The Epistle, however, is anonymous, and is not attributed to St. Paul by the most ancient historical testimony which has come down to us, nor is his authorship generally recognized by modern criticism. It is interesting, therefore, to enquire whether St. Chrysostom, in adopting the prevailing view of his time, did so on sufficient grounds.
The history of the matter is very curious. At the close of the second century Tertullian speaks positively and unhesitatingly of the Epistle to the Hebrews as written by Barnabas, the early and long-continued companion of St. Paul.  But there happened to be current in the ancient Church another epistle ascribed to Barnabas, and then commonly received as his, though generally considered spurious. The two epistles were so entirely unlike that no one could well receive them both as from the same author. The result was different in different parts of the Church. In the West, although the Epistle to the Hebrews had been used very largely by Clement of Rome, it came to be discredited altogether, and did not secure general recognition until the fourth century; it was then gradually acknowledged and attributed, at first doubtfully, but afterwards by common consent, to St. Paul. In the East, on the other hand, the Epistle itself was firmly accepted from the first, but with no certain tradition and much questioning in regard to its author. The suggestion of its Pauline authorship seems to have been made by Pantænus, the teacher of Clement of Alexandria, and a contemporary of Tertullian. We have his opinion, however, only at third hand, in a quotation preserved by Eusebius  from a lost work of Clement, and it is impossible to tell on what grounds he rested his opinion, or whether it was a mere personal speculation, like the reason he gives for the omission of the name of St. Paul in connection with the Epistle.
His disciple Clement adopted the suggestion not without hesitation. No one familiar with Greek, which was still the current language of the East, and especially of Alexandria, could fail to be struck by the extreme difference of style between this Epistle and those of St. Paul. Clement, therefore, conjectured that it might have been originally written by St. Paul in Hebrew and translated into Greek by St. Luke. This again is second-hand opinion preserved to us by Eusebius.  Nevertheless, in other works, which are still extant, he frequently cites the Epistle as St. Paul's.
Clement was succeeded in his catechetical office at Alexandria by Origen, a profound thinker and scholar. He was strongly impressed with the difference between the Greek of this and of the Pauline Epistles, and speaks of the matter in different parts of his voluminous works, sometimes suggesting the Clementine hypothesis, sometimes speaking of the variety of opinions and traditions on the subject, sometimes speaking of St. Luke or of Clement of Rome as the probable author, but summing up his perplexity (in language, quoted fully by Eusebius), by saying that who really was the author, God only knows. 
Thus far the question of authorship was evidently an open one on which everyone was free to hold his own opinion, or uncertainty of opinion. Tertullian speaks of the authorship of Barnabas simply as a fact, without an allusion to any doubt on the matter. But as the time went on, the attention of the masters of thought in the Church became more and more engrossed with doctrinal questions, while those of exegesis and criticism more and more lost their interest, especially in the East. In the West there is no trace of any reference of the authorship of the Epistle to St. Paul until the middle of the fourth century; but after this the opinion spread rapidly, and under the influence of Augustine, in the year 393 somewhat hesitatingly, but in 419 positively, the provincial council of Carthage reckoned it among the Pauline Epistles. Augustine himself, however, sometimes expressed himself doubtfully, and although it had now become customary to quote the Epistle as St. Paul's, yet scholars like Jerome, when distinctly treating of the question, express the old doubts and uncertainties of Origen. The assumption of the Pauline authorship was a convenience in maintaining the authority of the Epistle, and there being almost no one to call it in question, had come to be generally adopted in St. Chrysostom's time, and remained almost unquestioned until the revival of learning at the period of the Reformation. Since then, while still remaining a popular impression, it has come to be rejected by the great majority of careful students.
In this variety of opinion from the earliest times, and in the absence of any consistent external evidence, we are plainly left free to form our own conclusions from internal evidence. Among the great number of authors suggested by different writers, the only names entitled to especial consideration are those of St. Paul (Chrysostom, Augustine, and later writers generally until modern times, but at present the only scholar of weight is Hofmann), St. Luke (besides the views of ancients given above, Calvin, Ebrard, Döllinger, and to a certain extent Delitzsch), Clement of Rome (Erasmus, Reithmaier, Bisping), Silas (Mynster, Böhme, Godet), Apollos (Luther, Semler, De Wette, Tholuck, Bunsen, Kurtz, Farrar, De Pressensé, Bleek, Hilgenfeld, Lünemann, Alford), and Barnabas (Ullmann, Wieseler, Ritschl, Grau, Thiersch, Weiss, Renan, Keil). Of the three first we have genuine writings with which to make a comparison; of the three last -- assuming the spuriousness of the so-called Epistles of Barnabas -- nothing remains.
The supposition of the authorship of St. Paul, although so long carelessly held, seems almost forbidden by an expression in the Epistle itself. St. Paul was always most strenuous in asserting that he had received his apostleship and his knowledge of the truth "not of man, neither by man, but by Jesus Christ, and God the Father" ( Gal. i.1 ), while the author of this Epistle ranks himself among those who had received through the medium of others that Gospel "which at the first began to be spoken by the Lord, and was confirmed unto us by them that heard Him" ( ii.3 ). All attempts to weaken the force of this evidence by considering the passage as merely an instance of the rhetorical figure koinosis, in which the writer identifies himself with his readers, and thus attributes to himself what properly belongs only to them, have been unsuccessful. Delitzsch considers that if the Epistle were the joint work of St. Paul and St. Luke, in which the former only supplied the general course of thought, leaving its expression entirely to the latter, even this expression, so singularly like Luke i.1, 2 , might have been used; but this can only be by a practical surrender of the Pauline authorship. St. Paul everywhere lays such emphasis on the fact that his presentation of Christian truth was in no way whatever derived from man, but was from express divine instruction given to himself personally, that this passage must form a presumption against the Pauline authorship so strong as to be set aside only by clear and positive evidence. It has already appeared that there is no such external evidence; the internal will be examined below.
The authorship of Clement of Rome may also be set aside on two grounds: (1) That he quotes largely from this Epistle with the whole air of one citing from a higher authority to confirm his own teachings; and (2) that his own manner and style, as well as intellectual power, is so unlike as to make the supposition of a common authorship scarcely conceivable.
The early suggestion that the Epistle may have been written in Hebrew by St. Paul, more or less fully, and translated by St. Luke or St. Clement, or some other of his companions more or less paraphrastically, can find no favor with the modern scholar. If such a supposition is meant to leave the work essentially a translation, it encounters all the difficulties already mentioned against the Pauline authorship, and besides is opposed to abundant evidence that the work was originally written in Greek. "It abounds in compound words which are essentially Greek, which have no analogues in Aramaic or in Hebrew,"  and it contains paronomasia, entering into the thought, which could only be possible in Greek. If, on the other hand, it is meant to express merely some connection of St. Paul with the thought and line of argument of the Epistle, it really gives up the Pauline authorship, and even this thread of connection may be found in the sequel difficult to retain.
In favor of the authorship of St. Paul so far as the ideas and essential argument of the Epistle are concerned, Origen urges the beauty of the thoughts, and there must be some force in this argument, or the Epistle could hardly have been so long and so widely attributed to him. Perhaps it may be summed up in the words of an eminent and now departed divine,  "If the Epistle were not written by St. Paul, then we have the remarkable phenomenon that there were two men among the Christians of that age who were capable of writing it." The theory has also a certain primâ facie probability, and offers a convenient way of reconciling the conflict of the external evidence. But of course it cannot be accepted merely on these grounds.
At the outset, on a general view of the Epistle, every one must be struck with the marked difference in its construction from any of St. Paul's Epistles. The omission of his name at the beginning has been more or less satisfactorily accounted for from ancient times, but the reasons for this do not apply to the absence of any sort of salutation, "any heading or introductory thanksgiving," by which St. Paul always takes pains to conciliate his readers, and of which there was especial need if he were writing to Hebrews disposed to prejudice against him. On the contrary, after the manner of St. Mark in his Gospel, the writer strikes directly into his subject, without any sort of preface. Another striking feature of difference is, that St. Paul always keeps close to his argument until it is complete, and then adds practical exhortations founded upon it, while in our Epistle each short division of the argument is separated from that which follows by its appropriate practical application. This indicates quite a different habit of mind, and it is difficult to fancy such a severely logical reasoner as St. Paul thus pausing in the flow of his argument. The style of the Epistle is so markedly different from that of St. Paul that attention has been drawn to this point from the time of Origen down. The "rounded oratorical periods" of the Hebrews are very unlike the "unstudied, broken, abrupt phraseology" of St. Paul. This difference might, in part at least, be accounted for as the work of the translator; only in that case, the translator could have been neither St. Luke, whose style is clear and smooth enough, but not at all oratorical, nor Clement, whose style is very unlike.
When we come to details, there are two passages which have been thought to favor a Pauline authorship. There is a quotation in Heb. x.30 , which, it is alleged, agrees precisely with the same quotation in Rom. xii.19 , but differs from either the Hebrew or the Greek of Deut. xxxii.35. The A.V. makes a slight variation in language between Romans and Hebrews, but the Textus Receptus of the original is the same: "Vengeance is mine; I will recompense, saith the Lord." Now the LXX reads, "In the day of vengeance I will recompense"; the Hebrew, "mine [are] vengeance and recompense." If, however, we examine any critical text, we shall find that the clause "saith the Lord," is rejected as a gloss in this Epistle, while undisputed in Romans, thus constituting a difference between them. It is still true, however, that they both differ in the same way from the Hebrew and the LXX. This might be a difficulty were it not that the quotation as it is in this Epistle is found exactly in the Targum, and from that had probably passed into familiar use. Everywhere else the author of Hebrews quotes very closely from the LXX, and from that in what is known as its Alexandrine form, while St. Paul uses the Vatican text, quotes far more loosely, and often follows the Hebrew rather than the Greek.
The other passage really gives no clear indication at all, and as far as it goes, is rather at variance with Pauline authorship. In xii.23 the writer says, "Know ye that our brother Timothy is set at liberty; with whom, (if he come shortly,) I will see you." It is of course possible that Timothy may have been imprisoned, at Rome or elsewhere, when St. Paul was with him; but as far as we know the history of the two, it seems unlikely. The passage might quite as well have been written by almost any of the companions of St. Paul who were also associated with Timothy.
When now, enquiry is made as to the indications to be found in the choice of words and construction of sentences, there is certainly room for some difference of opinion. Delitzsch has endeavored throughout his commentary on this Epistle to show that there is such a striking similarity between it and the writings of St. Luke as to favor decidedly the view that it was written by him; Lünemann, on the other hand, in the introduction to his commentary, has collected the instances of Delitzsch and remarks upon them, "So soon as we separate therefrom that which is not exclusively peculiar to Luke and the Epistle to the Hebrews; so soon as we also put out of the account that which Luke has only taken up out of the sources employed by him, and cease to lay any weight upon isolated expressions and turns of discourse which were the common property either of the Greek language in general, or of the later Greek in particular, and are only accidentally present in Luke and the Epistle to the Hebrews, -- there is nothing whatever left of an actual affinity, such as must of necessity admit of being traced out between the works of the same author." The fact seems to be that there is between these two writers as compared with the other New Testament writers a certain similarity, not so much of particular words and constructions, as of the general cast, both of the phraseology and the structure of the sentences; but that this similarity arises, not from the identity of the writers, but from the fact that both wrote in somewhat better Greek than is found in the rest of the New Testament. The grammars of the New Testament Greek continually refer to the fact, that certain classical constructions are found only, or at least more frequently, in these writers than elsewhere. But this does not prove more than that the author of this Epistle, as might easily have been the case with several of the companions of St. Paul, like St. Luke, was more accustomed to classical Greek usage than most of the earliest Christian writers.
An examination of the vocabulary of this Epistle in comparison with that of St. Paul, St. Luke, and the other New Testament writers will throw some light upon the question. In another place  I have made such an examination with some care, and will here give a summary of its results. It is to be borne in mind that this Epistle is much shorter than the collective writings of St. Paul, or St. Luke, or of the other New Testament writers taken together. By a careful estimate of the actual length of these four groups it is found that, taking the longest as the standard, in order to determine the relative use of any word in them, it is necessary to multiply the number of its occurrences in St. Luke by 1.57, in St. Paul by 1.86, in Hebrews by 11.56. The results may in many instances prove fallacious. Any writer may use a word several times, even in a short passage, which he would not have used again had his writing been greatly extended; or he may not use a particular word once in twenty pages, when he will employ it several times in the twenty-first. Such facts must be borne in mind, but the above process seems to be the only means of making a comparative statement in figures; and when it is applied to a large number of words, and especially to whole groups of words which correspond to certain classes of ideas, the general result must have a decided bearing upon the question of authorship.
It has often been noticed that the number of words peculiar to any New Testament writer is an index of the number freely at his command. Peculiar words, it is true, are often required by peculiarity of subject, and may sometimes be what is called accidental. Still, when the number of them in any writer is unusually large, the fact has its value, and such words do abound in the writings of St. Luke and in the Epistle to the Hebrews above all others.  No great importance perhaps should be attached to this point; yet as it is often brought forward, the exact facts should be ascertained. Excluding words occurring only in quotations from the LXX (which can have no bearing upon the characteristics of the writer), and also excluding words which depend on doubtful readings, the number of words found in the New Testament only in the Gospel of St. Luke is 249, in the Acts 414, in both taken together 724; the similar number in the much shorter Epistle to the Hebrews is 147, while even the Apocalypse, with all its peculiar subjects and imagery, has but 116, and none of the other books (except Matthew 114) reach as high as 100. This suggests that the writer of this Epistle was like St. Luke in having at his command a peculiarly rich vocabulary. But if the facts be looked at in another way, and the comparative length of the various books taken into consideration, a different result is reached. St. Luke's Gospel has one peculiar word to every 9.76 lines; Acts, one to every 5.77; Hebrews, one to 4.45; but 1 Timothy has one to every three lines; 2 Timothy, one to 3.22; Titus, one to 2.97; James, one to 3.5; and so on with several of the shorter epistles. The result of such statistics appears to depend much upon how they are manipulated. Nevertheless, in no book of nearly equal, or of greater length, is the proportion so large as in this Epistle, except in the Acts. If the writings of various authors be taken collectively, --
St. Luke has 724 peculiar words = 1 to every 6.66 lines.
St. Paul has 777 peculiar words = 1 to every 5.25 lines.
Hebrews has 147 peculiar words = 1 to every 4.45 lines.
St. John has 244 peculiar words = 1 to every 13.46 lines.
All others taken together have 378 peculiar words = 1 to every 11.38 lines.
On the whole, then, the first impression of every reader is confirmed: St. Paul, St. Luke, and the author of Hebrews are alike distinguished from the other New Testament writers by the comparative richness of their vocabulary; yet, in view of the peculiar subjects treated in this Epistle, this fact has less significance than it might be entitled to under other circumstances.
Another question may be asked of the same kind. May not some indication of authorship be found in the number and character of the words common only to the Hebrews with St. Paul, with St. Luke, and with the other writers respectively? There are 34 words common to St. Luke and Hebrews, and found nowhere else; to St. Paul and Hebrews, 46; to all others and Hebrews, 28. Or, proportioning these numbers to the length of the several books, common to Luke and Hebrews, 53.5; to St. Paul and Hebrews, 85.56; to other writers and Hebrews, 28; or nearly twice as many common to Hebrews with St. Luke, and more than three times as many common to Hebrews with St. Paul, as there are common to Hebrews with other writers. This examination tends like the other, but much more strongly, to connect this Epistle both with St. Luke and St. Paul, but especially with the latter. It falls in with the vacillating opinion of Origen, already given, and with his report of the current traditions of his time.
But much more important than the mere numerical statement, is the character of some of these words, used in common by these writers and by no others. Most of them, indeed, have nothing characteristic, and many are used but once by each of the writers, and that apparently without any special design. There are several, however, worthy of more consideration. The noun katapausis and the verb katapauo , which might be expected to be common enough, are used only in Luke and Hebrews, the noun once in Luke, eight times in Hebrews; the verb 11 times in Luke, three times in Hebrews; or together, 12 times and 11 times. The noun metochos is used once in Luke, five times in Hebrews, and nowhere else, while the verb metecho is used five times by St. Paul, three times in Hebrews, and by no other writer. Diatithemai occurs three times in Luke, twice in Hebrews, and nowhere else. Sunantao is used four times in Luke, twice in Hebrews, and not elsewhere. The word for star in Greek has either form, a stron or a ster, and both are common in the LXX; but the former is used exclusively by St. Luke (three times) and also in Hebrews, where, however, it occurs but once; but a ster is used exclusively by all the other New Testament writers, by St. Paul three times, by others 21 times. On the other hand, e ndeiknumi occurs twice in Hebrews, nine times in St. Paul, and in no other writer. The verb euaresteo (occurring three times) is peculiar to Hebrews, as is also the adverb euaresteos, while the adjective euarestos occurs once in Hebrews, and seven times in St. Paul, being found nowhere else in the New Testament. The striking adverb e phapax, not found in any other New Testament writer, occurs three times in Hebrews and twice in St. Paul. The verb leitougeo with the nouns leitougia and leitourgos and the adjective leitourgikos, though common enough in the LXX, and apparently sufficiently often called for, are used in the New Testament only by St. Luke, St. Paul, and in Hebrews. The verb occurs once in each of them; leitourgia is used once by St. Luke, three times by St. Paul, twice in Hebrews; leitourgos three times by St. Paul, twice in Hebrews; while the adjective occurs only once in the last; i.e. taking the whole group together, it is employed twice by St. Luke, seven times by St. Paul, and six times in Hebrews, and never elsewhere. The much more important word mesites is used only in St. Paul and Hebrews, three times in each. The same is true of ho mologia, a word which might have been expected more frequently. There seems to be nothing peculiar about o neidismos which yet happens to be found only in St. Paul (three times) and in Hebrews (twice). The words paideia and paideutes also occur only in these writers, the former four times, the latter once in St. Paul; the former twice, the latter once in Hebrews; or together, five times and three times. We are surprised to find such a word as plerophoria only in these writers, in each of them twice. The remarkable word hu postasis, afterwards in another sense of so much importance theologically, is found only in these writers, in St. Paul three times, in Hebrews twice.
The results of this comparison have a positive value, unless they can be, at least in some good degree, paralleled by words common to Hebrews and the other New Testament writers. I do not find this to be the case. There seem to be but two words common only to Hebrews and to any of them occurring more than once in each. One of these is the purely accidental word he bdomos, used twice in Hebrews, and seven times elsewhere (five times in Revelation); and the other is the more important word baptismos (always in the plural = purifying ablutions) used twice each in Mark and Hebrews. Whatever value, therefore, there may be in this examination of common words, it is much increased by the almost entire absence of any such relation between this Epistle and the other writings of the New Testament. It certainly points, as far as it goes, to some sort of relation between the three writers, St. Luke, St. Paul, and the author of Hebrews, and especially between the two last.
We now turn to common words of wider range which yet have something in their usage tending to show the style of the writer. The verb e rchomai with its compounds a p -, e p -, e x -, eis -, kat -, par -, and pros -, is naturally more common in narrative. Making allowance for this, we are surprised at its relative frequency in Hebrews and infrequency in the Pauline Epistles, while the word is in such common use as to make this difference significant. The proportionate numbers are: Hebrews, 519; St. Luke, 656; St. Paul, 169; all others, 708. For the particular compound eiserchomai, the same numbers are: Hebrews, 196; St. Luke, 133; St. Paul, 7; all others, 91. While it is relatively much the most frequent in Hebrews, it is yet common in St. Luke, but almost entirely avoided by St. Paul.
Lambano with its compounds e pi -, para -, and hu po -, have a similar variable usage. They are all relatively much more frequent in Hebrews than elsewhere, less common in St. Luke, and still less so in St. Paul; taking the simple verb and its compounds separately, St. Luke alone uses that with hu po (four times actually, or relatively, six times), and almost entirely avoids that with para, and St. Paul, like the other writers, that with e pi ; while Hebrews uses them all (except hu po) with peculiar frequency. The proportionate numbers are: --
Hebrews. St. Luke. St. Paul. All Others.
e pi - 34....19......4....4
para - 139......1....20..25
hu po - -- ......6.... -- .. --
-- -- -- -- -- -- -- --
The verbs employed for request or prayer are numerous, and their employment by the different writers varies much. The following list of their relative frequency shows the principal facts: --
Hebrews. St. Luke. St. Paul. All Others.
a p -....... -- .........3....... -- ..... --
e x -....... -- .........2....... -- ..... --
e p -....... -- .........3....... -- ..... --
par -.....35.........6.........7..... --
pros -.... -- .........1....... -- .......2
deomai. -- .........8.........9.......1
e rotao....... -- .......31.........7.....37
euchomai -- .........3.........6.......2
-- -- -- -- -- -- -- --
While St. Matthew habitually designates heavenly things by the plural (gen. or dat.) of ouranos, and is somewhat followed by the other writers, the author of Hebrews and St. Paul employ these forms very little and are almost alone in availing themselves of the compound adjective e pouranios for the same purpose. St. Luke uses this word only once; Hebrews, six times; St. Paul, twelve times; all others, twice. On the other hand, the simple ouranios is not used at all in Hebrews and St. Paul, but occurs twice in St. Luke, and four times elsewhere.
The words lalein and legein are both common enough, and the distinction between them is well recognized. The point to be noticed is the frequency of their use relatively to each other. Hebrews uses them in the proportion of 1:2; St. Paul the same; St. Luke, 1:3½; all others, 1:7 nearly. St. Luke here varies considerably from Hebrews and St. Paul, but far less than the others.
The Hebraistic prosopon, so frequent in the LXX, is found but once in Hebrews, and then in an allusion to the LXX; but curiously occurs 27 times in St. Luke, as many in St. Paul, and in all others 22. So also rh e ma is a common enough word; but in its Hebraistic sense, corresponding to dabar, a thing, the subject-matter of speech or command, its use is confined to St. Luke, and it does not occur either in Hebrews or elsewhere. Both hu parcho and hu postrepho are favorite words with St. Luke. The former occurs in his writings 34 times, is not found at all in Hebrews; is used by St. Paul 11 times, and only four times elsewhere; the latter is used by St. Luke 31 times, and elsewhere only once each in Hebrews, St. Paul, and St. Mark. Katergazomai is a Pauline word (21 times), never used in Hebrews, and but three times elsewhere (Jas.2, Pet.1). On the other hand, the use of the comparatives kreisson and pleion, with the superlative pleistos, is far more common in Hebrews. The comparative numbers are: for kreisson, Hebrews, 150; St. Luke, ; St. Paul, 7 (he also used the adverb kreisson once); all others, 2. For pleion pleistos, Heb., 46; St. Luke, 42; St. Paul, 19; all others, 17.
I do not recall any other words of this kind, the usage of which affects our enquiry. Such inferences as may be drawn from this examination are somewhat contradictory. They certainly do not point to the author of this Epistle as either St. Paul or St. Luke, as they might be expected to do if such were the fact. There are some striking similarities of diction; but the differences are, at least, quite as important.
It is time now to turn to those adverbs, particles and prepositions, which bring out the grammatical form in which a writer is accustomed to clothe his ideas. But before speaking of these, mention must be made of one grammatical form peculiarly characteristic of the nicety and subtlety of thought of the classic Greek writers -- the optative mood. This subject has been investigated by Dr. Harman with great care. He finds that this mode is used in the whole New Testament 66 times, 32 of them in the Pauline Epistles, 28 in the writings of St. Luke, once in Hebrews, and five times in all other writers. "In nearly all the cases in which the optative occurs in the New Testament it is used to express a wish or prayer, except in the writings of Luke." "In the Epistle to the Hebrews we find one instance of the optative, katartisai, May God make you perfect' ( xiii.21 ). This is presumptive proof that an Alexandrian did not write this Epistle, as it is not likely that the use of this mode in but one instance would have satisfied his fine Greek taste." 
The usage of the particles, adverbs and prepositions, require so much detail that only a summary can here be given with a reference to the paper on the "Language of the Epistles to the Hebrews," already mentioned.
The particles men and de would naturally be more common in narrative; but as between Hebrews, St. Paul, and the other Epistles, they are relatively most frequent in Hebrews. The same is true of the conjunction te, which is more common in St. Luke and St. Paul than in other New Testament writers. The adversative a lla is common enough everywhere, not even the shortest epistle being without more than one instance of its use. St. Luke and Hebrews employ it very much more seldom; but again, St. Luke uses it far less than Hebrews. The three writers, St. Luke, St. Paul, and the author of Hebrews, are distinguished from the other writers by the (comparatively) sparing use of a n and e a n , but of the three, St. Luke employs it most, and St. Paul least. Hebrews alone uses e a nper, but never e a n me, which is employed in St. Luke with moderate freedom, by St. Paul twice as often, and still more frequently by the other writers. In the case of dio, St. Paul uses it relatively only half as often as Hebrews, but three times as often as St. Luke, and the last more than twice as often as the other writers. In the use of dioti, a much less common word, there is a less difference, but still a marked one and in the same order.
The pronouns of the first and second person are used, as might be expected, most abundantly by St. Paul; but Hebrews is singularly shy of them. This fact has been noticed and an explanation offered on the ground that the work has more the character of a treatise than of a personal epistle; but this explains too much, since these personal pronouns, though relatively infrequent, are still very common in our epistle. The author was not disposed to bring forward the personality of either himself or his hearers. St. Paul, on the contrary, used these pronouns more than twice as often as our author, and indeed far more frequently than any other New Testament writer.
The case of the third person of the pronoun is peculiar, since its frequent redundant use is one of the marked characteristics of the New Testament diction; yet St. Paul uses it less than half as often as Hebrews. The same is true of the demonstrative e keinos ; but in the use of the reflective he autos, St. Paul is largely in excess. The difference in the use of these pronouns between Hebrews and St. Luke is not very great; but, in regard to the two first particularly, the difference from the other writers is marked, and St. Paul's usage of all of them is very different. On the other hand, in the use of houtos, Hebrews is strongly separated from St. Luke and less so from St. Paul.
The words a llelon, a llos, he teros, tis (interrog.), and tis (indef.) have marked peculiarities in their frequency of employment by the different writers, but it is enough to instance here a llos, employed oftener by St. Paul than by St. Luke and Hebrews put together, and yet by the other writers collectively twice as often as by him; and he teros used with exactly the same frequency by St. Luke and St. Paul, less than one quarter as often by other writers, but far oftener in Hebrews than in any of them. So also with he kastos, toioutos, and tosoutos : they are all used with exactly the same frequency in Hebrews; but while the first is used by St. Paul much oftener, the last is used only one-fourteenth as often. The usage of St. Luke is markedly different from that of either.
Further and more detailed examination of words of this class would be out of place here. Suffice it to say that such an examination shows a marked individuality in the usage of the several writers; and it is to be remembered that Hebrews, Acts, and the latter Epistles of St. Paul must have been written with no great interval of time between them.
The same things are true in whatever way we test the forms of expression of these writers. If we take the particle ei with its various combinations ei kai, ei me, eige, ei de me, ei me te, ei per, ei pos, ei te, ei tis, we shall find that only the first two of these combinations occur in Hebrews at all, and those only once each, while all of them are found in the Pauline Epistles, and all but three of them in the writings of St. Luke. The whole group together is used more than twice as often by St. Paul as by any other writer.
Adverbs of space are very sparingly used by St. Paul, with an approach to equality between Hebrews and St. Luke, but twice as often by other writers. In the various particles and adverbs of negation, there do not seem to be, on the whole, very noticeable peculiarities, although uedeis and mede occur but once each in Hebrews, while the former is everywhere else common, and the latter also in St. Paul and other writers, though less frequent in St. Luke. But the word choris is frequent enough in Hebrews to be considered characteristic, is used far less by St. Paul, only once by St. Luke, and comparatively seldom by other writers. Palin is quite rare in St. Luke, equally common in Hebrews and St. Paul, and a little less so in the other writers. The use of e is very rare in Hebrews whether as a disjunctive conjunction, or as a term of comparison. In the latter sense it occurs but once, and in the former only four times, two of which are in quotations from the LXX, and a third in a more than doubtful reading. In all the other New Testament writers it is very common, but most of all in St. Paul. In the use of mallon, however, though St. Paul still exceeds, Hebrews most nearly approaches his usage. The causal e pei is five times as frequent in Hebrews as in St. Paul, and yet four times more frequent in his writings than in the others, and is still less common in St. Luke than in them. The word pos is used but once in Hebrews (interrogatively), while it is common enough everywhere else. Both hi na and ho ti are used very often by St. Paul and the other writers, and much more sparingly in Hebrews; but St. Luke uses hi na much less than half as often as Hebrews, while he employs ho ti much more than half as often again. The un-Attic particle kathos, the adverbs houtos, ho sei, ho sper, and the conjunction ho ste, are all appropriate to trains of reasoning, but their usage in the different writers, particularly in the three we are especially considering, is very various. The same may be said of the prepositions, among which sun is never used in Hebrews (except in composition), while it is employed much oftener by St. Luke than by St. Paul or any other writer.
It may be thought that all this examination -- still tedious, though much condensed -- is not worth the trouble. It goes to show, what has always been noticed by every reader, that the style of this Epistle is unlike that of St. Paul; but if it show, as it seems to do quite as clearly, that it is unlike that of St. Luke as well, something has been gained. It makes it at least improbable that St. Luke wrote the Epistle to give expression to the ideas of St. Paul.
It remains to examine some words of another class. There are many words and groups of words so peculiarly appropriated to certain ideas or shades of thought, that the use or non-use of them becomes a fair index of the habitual tone of thought of the writer. If he use them frequently, the phase of truth which they represent must have been prominent in his mind; or if he seldom employ them, then that aspect of truth was not the predominant one from his point of view. Such words or groups of words may be of different degrees of importance; but even those of inferior significance help to complete the picture of the writer's mental habits, and it is therefore well to examine all that are at all characteristic. As the force of the evidence from these words can only be brought out by a more careful examination of them, I venture to copy some pages of the paper referred to above.
The group a gapao, a gape, and a gapetos is noteworthy. They are very common in the Pauline writings, but are rare both in St. Luke and in Hebrews. In fact only one of them, a gapetos, occurs at all in the Acts, and none of them are ever used by St. Luke except in recording the words of others. So also of the Hebrews. Of the five instances of their use, two are in quotations from the LXX. They are common enough in other writers, but are special favorites of St. John. Of the 154 instances in "other writers," 109 are in St. John, so that the words may be called Pauline and Johannean. Their rarity in St. Luke and Hebrews may be partly explained by the fact that a gape is an exclusively biblical word, and that a gapao also is used in a higher sense in the sacred than in profane writings. Still they were common words in the Christian community, and they mark a distinction in thought between St. Luke and Hebrews on the one side, and St. Paul and the rest of the new Testament on the other. The actual number of instances of their use is: Hebrews, 5 times; St. Luke, 15; St. Paul, 135; all others (John, 109), 154; but if we exclude from the enumeration all quotations from the LXX, and all record of the words of others, the numbers become: Hebrews, 3; St. Luke, ; St. Paul, 132; all others (John, 43), 87. The comparison is too obvious to call for proportionate numbers. As an appendix to this group it may be mentioned that phileo never occurs in Hebrews, is used only twice by St. Luke, twice by St. Paul, and 21 times elsewhere, 15 of which are in St. John.
A word especially appropriate to Hebrews, and one which might have been expected there very often, ha giasmos, occurs but once, while it is used eight times by St. Paul, and is not found in the rest of the New Testament. On the other hand, haima, which we might have expected frequently in St. Paul as well as in Hebrews, is very common in the latter and not at all so in the former. Proportionate numbers are: Hebrews, 231 times; St. Luke, 31; St. Paul, 24; all others, 44. Here, from the nature of the writings, we may not be surprised at the commonness of the word in Hebrews; but its comparative rarity in St. Paul is remarkable. His subjects led to it, and had it come to his mind as readily as it did to that of the author of Hebrews, it must have occurred in his writings much oftener.
The group of words, a letheia, a lethes, a lethinos, a lethos, and a letheuo, which we are accustomed to consider peculiarly Johannean, is also very frequent in St. Paul, but comparatively rare in St. Luke and Hebrews. The actual numbers are: Hebrews, 4 (a lethinos three times, a letheia once); St. Luke, 12; St. Paul, 55; all others, 114 (of which St. John, 95).
Of the group a stheneia, a stheneo, a sthenema, and a sthenes, only two occur in Hebrews -- a stheneia (four times) and a sthenes (once). The actual occurrences of the whole together are: Hebrews, 5 times; St. Luke, 15; St. Paul, 43; all others, 21; or proportionately, Hebrews, 58; St. Luke, 23; St. Paul, 90; all others, 21. This is evidently an especially Pauline class of words.
The words of opposite signification, bebaios, bebaioo, bebaiosis, -- very infrequent in the LXX, -- do not occur at all in St. Luke, and are relatively far more frequent in Hebrews than anywhere else. The proportionate numbers are: Hebrews, 92 times; St. Luke, ; St. Paul, 15; elsewhere, 3. Evidently St. Paul preferred to dwell upon weakness, the author of Hebrews upon strength.
There is a similar contrast between the verb e lpizo and the noun e lpis on the one hand, and the verb e pangellomai and the noun e pangelia on the other. Kleronomeo, kleronomia, and kleronomos are most common in Hebrews. The proportionate numbers are: --
Hebrews. St. Luke. St. Paul. All Others.
e lpizo and e lpis.69...................20.................100.................10
e pangellomai and e pangelia..208...................16...................57...................9
kleronomeo, kleronomia, kleronomos..104...................11...................33...................8
-- -- -- -- -- -- -- --
It is plain that while the author of Hebrews dwelt much more upon the brightness of the future than any other writer, he preferred to speak of it in the light of promise and of inheritance, while it rested in St. Paul's mind more as a hope. This is the more noteworthy because the ideas of sonship and of adoption are very common in St. Paul. He alone uses the word huiothesia five times.
The words he mera and semeron are curiously infrequent in a writer of the present urgency of St. Paul, and are relatively most common in St. Luke and Hebrews, but most so in the last. In proportion the numbers are: Hebrews, 278; St. Luke, 207; St. Paul, 100; all others, 153. That is, Hebrews uses them nearly three times as often as St. Paul.
The names for God and for our Lord are used by the various writers with much difference, and with an evident preference in each of them for his own accustomed word. The proportionate numbers (which can take no note of periphrases) are as follows: --
Hebrews. St. Luke. St. Paul. All Others.
In all cases St. Paul uses these words most freely (about twice as often as anybody else), except that in the case of ' Iesous he is exceeded by "other writers" as a result of the large amount of narrative contained in them. Theos is used in Hebrews next in frequency to St. Paul, but with a long interval between them, and very much more often than elsewhere. Kurios is used least frequently in Hebrews, while ' Iesous and Christos are employed there, one with exactly the same frequency as the other, though St. Luke, St. Paul, and the other writers employ them very unequally, one preferring one and another the other. The use of these words is so much a matter of habit, habit alike of writing and of mode of thinking, that these go far to differentiate the writers.
Kerugma, kerux, and kerusso are none of them ever used in Hebrews. For the others proportionate numbers are: St. Luke, 27 times; St. Paul, 35; all others, 20.
The group kauchaomai, kauchema, and kauchesis is almost exclusively Pauline, occurring in his writings 58 times, while it is nearly absent from Hebrews, only kauchema being used, and that but once. These words do not occur in the other New Testament writers except three times in James. So also logizomai and makrothumia are especially Pauline. They each occur only once in Hebrews. The first is found twice, the second not at all in St. Luke; but logizomai occurs 34 times in St. Paul, four times in other writers, while makrothumia is used by St. Paul ten times, and only three times elsewhere.
Manthano is used 16 times by St. Paul, only once each by Hebrews and St. Luke, and seven times elsewhere. Parakaleo and paraklesis are much more frequent in St. Paul's writings than elsewhere, but in this case he is more nearly approached by Hebrews than by others -- yet with a great difference. Proportionate numbers are: Hebrews, 81; St. Luke, 54; St. Paul, 137; all others, 22. The word proseuche occurs in Hebrews but once, and proseuchomai not at all. This is a noteworthy omission in our epistle, although it is also true that they are not used by St. John, except proseuche three times in Revelation. The two words are found in St. Luke and St. Paul each 33 times, and in the other writers 43 times. The words sarx and sarkikos are favorites of St. Paul. They occur seven times in Hebrews, six times in St. Luke, 102 times in St. Paul, and 46 times (of which one-half are in St. John) elsewhere. Proportionately Hebrews uses them about two-thirds as often as St. Paul, and nearly twice as often as all other writers together. The group phroneo, phronema, and phronesis is characteristic. None of them are found at all in Hebrews, and they occur but twice in St. Luke, and twice in the other writers (Matthew, 1, Mark, 1, in parallel passages), both in the record of the words of others; but St. Paul uses them 31 times. (He uses phronema, however, only in Romans -- four times.) A word used in a figurative sense especially characteristic of St. John (31 times), phos, never occurs in Hebrews. It is used 16 times by St. Luke, 12 times by St. Paul, and ten times by other writers. It is more or less used by every New Testament writer except the author of Hebrews, and St. Jude in his very short epistle. Chairo is also used by every other writer (St. Luke, 19 times; St. Paul, 27 times), except Hebrews and St. Jude. St. Paul greatly delights in the word charis, and in the idea conveyed by it; he never wrote an epistle without it, and uses it 101 times. In Hebrews it is found eight times, in St. Luke 24, and in all others 22, not occurring in the first two Gospels.
The foregoing list is somewhat long of words characteristic of phases of thought which are especially favorite with St. Paul, and either wholly unused or much less frequently employed in Hebrews. A corresponding list may be made of other words especially common in Hebrews, but less used by St. Luke and St. Paul. Before going to this, however, a few words are to be considered which in their frequency of usage are characteristic of all three, or of two of these writers as distinguished from others, although with some differences between them.
Most prominent in this latter class is nomos, which we are accustomed to think especially Pauline. It is indeed used much oftener by St. Paul than by any other writer, yet it also occurs in Hebrews with a frequency distinguishing that Epistle from any other writing. The proportionate numbers are: Hebrews, 162; St. Luke, 44; St. Paul, 227; all others, 33. Nevertheless, the assimilation here is more apparent than real; for St. Paul employs it chiefly of a method of salvation, while it refers in Hebrews mostly to a definite collection of statutes. In the same way pistis is usually regarded as a characteristically Pauline word. It is relatively much more common in Hebrews; for the proportionate numbers are: Hebrews, 369; St. Luke, 42; St. Paul, 262; all others, 43. But here also there is a shade of distinction in the force of the word as used by the two writers; St. Paul's pistis is reliance upon Christ as the means of salvation in opposition to the law and the works of the law, while in the Hebrews it is only a general reliance on God's grace and promises. Of course, it is not denied that St. Paul sometimes uses a word, so common, in such varied shades of meaning, in a more general way as in 1 Cor. xiii.2, 13; xvi.13; 2 Cor. v.7 , etc.; but the distinction in the shade of meaning between his habitual employment of the word, and that common in the Epistle to the Hebrews is easily recognized. In this connection pisteuo must be mentioned, though belonging in the former category. The proportionate instances of its use are: Hebrews, 23; St. Luke, 76; St. Paul, 100; all others (of which St. John, 99), 143. It is therefore a comparatively rare word in Hebrews. The adjective pistos, which ought perhaps hardly to be considered in this connection, is used proportionately, in Hebrews, 58 times; St. Luke, 6; St. Paul, 61; all others, 23. Peitho somewhat associates the three writers together, although most frequent in Hebrews, occurring proportionately, in Hebrews, 58 times; St. Luke, 33; St. Paul, 43; all others, 6. Suneideo and suneidesis are used proportionately: Hebrews, 58; St. Luke, 8 (all in Acts); St. Paul, 37 (but suneideo only once); all others, 4. The word soter, though used by St. Luke four times, St. Paul 12, and by others eight times, never occurs in Hebrews; but this is not remarkable, as it is not found in the much larger books of Matthew, Mark, and Revelation; moreover, it should be taken in connection with soteria and soterion, which also do not occur in the first two Gospels, but are found in Hebrews 7 times, St. Luke 13, St. Paul 19, and all others 9 times. This would give them a relatively greater frequency in Hebrews; but they are also common words in St. Luke and St. Paul. The word psuche, while a little more common relatively in Hebrews, is yet frequent enough in St. Luke and other writers, though not a favorite with St. Paul. Proportionate numbers are: Hebrews, 69; St. Luke, 57; St. Paul, 29; all others, 54.
This leads to the third class of words -- those which, embodying certain sets of ideas, are characteristic of Hebrews in distinction from other writers, especially St. Luke and St. Paul. One of these is the idea of witness, expressed by martur, martus, marturia, marturion, martureo, and marturomai. This group of words is especially common in Hebrews, and much less frequent in St. Luke and St. Paul. It is also very common in St. John. The proportionate numbers are: Hebrews, 127; St. Luke, 45; St. Paul, 50; all others (of which St. John, 80), 96. The word taxis is so naturally called for in the argument of Hebrews that there is nothing remarkable in its occurring there seven times, while in all the rest of the New Testament it is found but three times (St. Paul twice, St. Luke once). The perfection and finality of Christian truth as set forth in this epistle comes out in the frequency of the use of these words as clearly as in its general scope; it is difficult to suppose that the Epistle to the Galatians, e.g., which does not contain any of these words, could have been written by the same author.
But by far the most important word in this connection is hi ereus, with its various derivatives, a rchiereus, hi erateia, hi erateuma, and a rchieratikos. The last two of these are of little importance, as hi erateuma occurs only twice, in St. Peter, and a rchieratikos only once, in Acts; also hi erateia occurs only once each in St. Luke and Hebrews, and nowhere else. Altogether, hi ereus and its compounds and derivatives occur 159 times, but are never once used by St. Paul.  The actual numbers are, for hi ereus : Hebrews, 14 times; St. Luke, 9; St. Paul, ; all others, 9; for a rchiereus, Hebrews, 17; St. Luke, 37; St. Paul, ; the other Gospels, 68, but never elsewhere. This is a remarkable fact. In view of St. Paul's arguments in the epistles to the Romans and to the Galatians, and in view of the frequency and emphasis with which he insists in all his Epistles, upon the sacrificial character of Christ's death, it seems to show that his mind was so absorbed in dwelling upon the value and power of the sacrifice that he was not in the habit of thinking or speaking of Christ as also Himself the Sacrificer. Redemption came to his thought through the medium of the Victim by whom it was obtained, but not through that of the Priest who offered the Victim. This is the more striking from the fact that he often speaks of Christ as giving Himself, offering Himself, and the like; but always for the purpose of bringing out the voluntariness and the love of the act, and never with any allusion to its priestly character. The line of reasoning in the Epistle to the Hebrews was thus quite foreign to the habitual thought of St. Paul. Such similarity of language to his acknowledged writings as exists must be accounted for in some other way.
On looking back over these various words, with their difference of usage, it is plain that they are not perfectly of accord in their indications. This was to be expected. I have endeavored, in this part of the examination, to select only words characteristic of thought, and to note every word of this kind in regard to which there is any considerable difference of usage; yet so many words are used by every writer accidentally, as it were, and not because they are characteristic, that much allowance is to be made. Still, the investigation seems to me to afford a sufficient basis for some probable conclusions. The Epistle contains both style-words and thought-words, characteristic alike of St. Luke and St. Paul, sometimes of one, sometimes of the other, sometimes of both; and these must be taken into account in any theory of the authorship. But they are not more than might be expected in any writer belonging among the companions of a leader of such magnetism and power as St. Paul. I see nothing in them to prove, hardly even to suggest, actual authorship. On the other hand, there are many words and groups of words expressing ideas very prominently in the mind of the author of this Epistle, which must have appeared also in the writings of St. Paul had the thoughts of this Epistle been derived from him, but which are not found there. Of course, no man expresses all his ideas in any one epistle, nor the same ideas in every one he writes; but the difference here is more radical. As one mind now is affected by one, and another by another of the various aspects of Christian truth, so the differences here go to show that the mind of the author of Epistle to the Hebrews was not affected in the same way as St. Paul; for Hebrews is scarcely more unlike the Epistles in which St. Paul addressed believing Jews than the speeches recorded in Acts xiii., xxii., and xxviii., in which he spoke to his still unbelieving countrymen. This leaves us free to accept the author's own statement, that instead of being, like St. Paul, one who had received his apostleship "not of man, neither by man, but by Jesus Christ, and God the Father" ( Gal. i.1 ), he was of that number who had received through the medium of others that Gospel "which at the first began to be spoken by the Lord, and was confirmed unto us by them that heard Him" ( Heb. ii.3 ).
It thus appears that neither are the thoughts of this Epistle Pauline, nor is its language that of St. Luke. It may be well to say a few words in conclusion as to the person to whom such facts as we have point as the probable author.
It is plain from what has been said, as well as from the common consent of students, that the author must be looked for among those companions of St. Paul who, through prolonged intercourse, were likely to have their modes of expression somewhat affected by his language. The number of these is considerable, and after so many ages of uncertainty, beginning with the earliest discussion of the subject, it is not likely that the right one can ever be pointed out with certainty. Many modern critics have selected Apollos as the most probable author, chiefly because of the facts recorded of him in Acts xviii.24-28 , that he "was born at Alexandria, an eloquent man and mighty in the Scriptures," and that after receiving further instruction from Aquila and Priscilla, "he mightily convinced the Jews, and that publicly, showing by the Scriptures that Jesus was Christ." He was certainly personally known to St. Paul ( 1 Cor. xvi.12 ), although of the length of time they may have been together we have no information. His being an Alexandrian is thought to explain what some are pleased to consider an Alexandrian tone of thought in the Epistle, and also the fact that its quotations are from the LXX, and accord rather with its Alexandrian than its Vatican recension. The force of the last point is not obvious. In the meagreness of our knowledge of the original LXX, it appears probable that the so-called Alexandrian recension was the one generally current in the Levant, and therefore that this indication, whatever it may be worth, simply points to an Oriental author. And so also whatever there may be of an Alexandrian caste of thought in the Epistle only indicates some one familiar with Jewish-Alexandrian literature, and this would include almost every educated Jew living in the Levant.  At all events, neither of these considerations seemed to have occurred to any of those early Alexandrian scholars, Pantænus, Clement, or Origen, who all speak of the authorship, the last at some length and with discrimination. The theory of Apollos' authorship has, however, this great advantage: that no line of his remains to compare with our Epistle. It has also these disadvantages: that it never occurred to any ancient author, but was first suggested by Luther; that there is no evidence of any prolonged personal intercourse between him and St. Paul; and that there is nothing to connect him with any especial interest in, or familiarity with, the Jewish ritual and temple beyond the simple fact that he was a Jew, as was also almost every other writer who has ever been suggested. The non-use of the optative is also strongly against the authorship of the Alexandrian Apollos. Moreover, it is clear from such passages as vii.12; x.32-36; xiii.7, 17-19, 23-25 , that this Epistle was addressed to some particular community, a fact now generally recognized, and that the author was personally and favorably known to his readers. There is a difference of opinion in regard to the locality of that community; but if, as seems altogether probable, it was Palestinian, we have no reason to suppose that Apollos was ever known to them; and although this evidence is only negative, it suggests looking for some other names positively in accord with it.
Of the other names suggested in ancient and modern times St. Luke and St. Clement of Rome seem to be sufficiently excluded by a comparison of the Epistle with their acknowledged writings; the former also by the probability that he was a Gentile, the latter by the very use he makes of the Epistle, apparently as quoting words of another. 
Silas has also been suggested as a possible author. Of him we know even less than of Apollos. He was a prophet in the early Church at Jerusalem ( Acts xv.32 ), and was the companion of St. Paul on his second missionary journey and subsequently in his labors at Corinth, and was also associated with the work of St. Peter ( 1 Pet. v.12 ). In all this there is nothing to mark him out as the one likely to have written this Epistle beyond several others of the companions of St. Paul. The only point which really gives plausibility to the suggestion of his authorship is the fact that he was much associated with Timothy ( 1 Thess. i.1; 2 Thess. i.1; 2 Cor. i.19 ), and this may explain the reference to Timothy in Hebrews xiii.23. On that ground the suggestion of his name might be adopted if there were not much more to be said in favor of another, and also if there were not the same very serious objection as in the case of Apollos -- that he was never so much as named in all antiquity.
There is a person, however, to whose authorship one of the very earliest witnesses, Tertullian, as already noted, positively and unhesitatingly testifies, -- Barnabas.  He has the same advantage with Apollos in having transmitted to us no writing with which to institute a comparison (the spuriousness of the epistle attributed to him being admitted,)  and in having been a Hellenistic Jew, likely to have written somewhat better Greek than St. Paul. His birthplace also was in the Levant, in Cyprus, where he could have had the full benefit of Alexandrian literature. Being at Jerusalem he became one of the very early converts to Christianity, long before St. Paul, and he was a man of property and benevolence; for although a Cypriote, he had land in Jerusalem and sold it to relieve the necessities of the early Christian community ( Acts iv.36, 37 ). He must have been known from the first very generally in the Hebrew-Christian community, and he must have been endeared to them, not merely by this act of benevolence, but by that kindly sympathy which led to his surname, "Son of consolation." A very few writers, indeed, have identified him with "Joseph called Barsabas, who was surname Justus," of Acts i.23 , and this is countenanced by the Codex Bezæ and the Æthiopic reading barnaban ; and in this case he must have been an original disciple, and would be excluded by the language of Heb. ii.3. But there seems to be no ground for the identification. In Acts iv.36 the language implies that Barnabas is there spoken of for the first time, the names themselves are different, and Barsabas was known by the surname of Justus, which does not appear to have been ever given to Barnabas. He is next heard of as bringing Saul, of whom all were afraid, to the apostles, and telling the story of his conversion ( Acts ix.26, 27 ), showing at once the position he occupied and his own moral courage. When tidings of the conversion of many Gentiles at Antioch came to the Church at Jerusalem, they sent forth Barnabas to take charge of the matter, and by his labors "much people was added to the Lord." The work growing too great for him, he sought out Saul at Tarsus and brought him to his assistance ( Acts xi.25, 26 ). Then after a year, the Church at Antioch sent Barnabas and Saul to carry their alms to the Church at Jerusalem. Having returned to Antioch, they were divinely selected to go forth upon a wider missionary work, in the course of which they visited "Lystra and Derbe," when probably the young Timothy received his first knowledge of Christianity. On St. Paul's second visit to these cities he is spoken of as already "a disciple." Barnabas must, therefore, have known him from the very beginning of his Christian life, and it is, therefore, entirely natural that he should speak of him in the way recorded in Heb. xiii.23. After Barnabas and Paul returned from this, when disputes arose between the Jews and Gentiles, they were sent to Jerusalem together, and having obtained a favorable hearing before the Council again returned to Antioch. Here are years of closest companionship between Barnabas and St. Paul, during all the earlier part of which Barnabas appears as the leader, Paul as the assistant. They had often stood together in the synagogue to tell to their fellow-countrymen the story of the cross, and probably had often discussed with one another the numerous Jewish converts. Barnabas must have been a man of dignity, for when the people of Lystra took them for gods, they selected Barnabas as Jupiter ( Acts xiv.11, 12 ). The companionship was broken up at the entrance upon another missionary journey, by a difference of opinion about taking Mark with them. In this case Barnabas, although doubtless influenced by his kinship, appears to have been the better judge of character, since at a later day St. Paul writes from Rome to Timothy, "Take Mark and bring him with thee; for he is profitable to me in the minis try" ( 2 Tim. iv.11 ); but however this may be, Barnabas showed in the matter independence and determination. He is called by the name of "Apostle" ( Acts xiv.14 ), and altogether held such a position in the Christian community as would make his writing such an Epistle a proper act. In all that is related of him there is but one faulty act, and even this points him out as especially interested in the Hebrews. When St. Peter behaved so ill at Antioch and received the sharp reproof of St. Paul, in his account of the matter St. Paul says, "the other Jews dissembled likewise with him"; and adds as evidence of the strength and danger of the defection, "insomuch that Barnabas also was carried away with their dissimulation" ( Gal. ii.13, 14 ). Barnabas then was not only a Jew by birth, but had strong sympathies with his race.
More than this: he was a Levite. The particular line of argument adopted in the main part of the Epistle to the Hebrews is one which would have occurred to few, and scarcely to any who was not familiar with the temple ritual. There is no evidence that this was the case with Apollos; but with Barnabas the temple service was a matter of professional duty, as well as the prompting of his devout heart. Indeed, an objection to the authorship of Barnabas has been based on this very point; -- it is said that the author does not show that nicely accurate precision in his statements which might be expected from one personally familiar with the temple. The points referred to admit of easy explanation on other grounds; but were they better taken, considering that the service of the Levites was altogether subordinate to that of the priests, and did not lead them into the naos itself, the objection seems hypercritical. But one of the actual duties of the Levites, and a very prominent one, was that of chanting in the Levitical choirs in the courts of the temple. This would have led to a special familiarity with the Psalms. Now it is a curious fact that about one-half of all the quotations from the Old Testament in the Epistle to the Hebrews are taken from the Psalms, and that the author cites that book, relatively, nearly four times as often as St. Paul, and eight times as often as St. Luke or the other writers. This fact is at once explained by the supposition that the author of the Epistle was a Levite.  It is not unlikely that when that "great company of the priests were obedient to the faith" ( Acts vi.7 ), Barnabas, as one of their attending Levites, was influenced by their example and with them accepted the faith of Christ.
The only important objection urged against the authorship of Barnabas is, that since the time of Tertullian until recently, there has never been any considerable weight of opinion in its favor. But this is accounted for by the almost universal acceptance in the meantime of the spurious Epistle of Barnabas as his genuine work. The two certainly could not have been written by the same person. The fact, however, that the spurious Epistle was attributed to him may be an indication of a belief that he had left to the Church some legacy of written teaching. Since that Epistle has been found not to be his, and is probably of a somewhat later day, there remains nothing to hinder the belief that the devout Levite of Cyprus, the early convert to Christianity while still in strong sympathy with the Christian Jews, the man of benevolence and wealth, and therefore probably of education, by birth the appointed servant of the temple, the man of independence and dignity, and yet of such tender sympathy as to be surnamed "Son of consolation," the long and intimate companion of St. Paul, and for years in the position of his superior, -- there is nothing to hinder the acceptance of the early ecclesiastical statement that he was also the author of the Epistle to the Hebrews.
 Tertull. De Pud. c. 20, Ed. Migne, 1021. Exstat enim et Barnabae titulus ad Hebraeos, adeo satis auctoritatis viro [viri], ut quem Paulus juxta se constituerit in abstinentiae tenore :...[1 Corinthians 9:6].... Et utique receptior apud Ecclesias Epistola Barnabae illo apocrypho Pastore moechorum. Monens itaque discipulos, omissis omnibus initiis, ad perfectionem magis tendere,...[After quoting Hebrews 6:4-8, he goes on] Hoc qui ab Apostolis didicit et cum Apostolis docuit, etc.  Eusebius' Eccl. Hist. vi. 14 (Crusé's translation, p. 213). "But now, as the blessed presbyter used to say, since the Lord who was the Apostle of the Almighty, was sent to the Hebrews, Paul by reason of his inferiority, as if sent to the Gentiles, did not subscribe himself an apostle of the Hebrews; both out of reverence for the Lord, and because he wrote of his abundance to the Hebrews, as a herald and apostle of the Gentiles.'"  Ibid . The Epistle to the Hebrews he asserts was written by Paul to the Hebrews in the Hebrew tongue, but that it was carefully translated by Luke and published among the Greeks. Therefore one finds the same character of style and of phraseology in the Epistle as in the Acts. "But it is probable that the title, Paul the Apostle, was not prefixed to it. For as he wrote to the Hebrews, who had imbibed prejudices against him and suspected him, he wisely guards against diverting them from the perusal by giving his name."  Eusebius' Eccl. Hist. vi. 25. Extended quotations from the various writers above referred to, and from many others, may be found in almost any of the innumerable treatises on the subject, and are given with especial fullness and clearness in Alford's Prolegomena.  Godet in The Expositor, April, 1888, p. 262.  Bp. George Burgess.  Journal of the Soc. of Bibl. Literature and Exegesis for June, 1887, pp. 1-27.  See Thayer's Grimm's N.T. Lexicon, Appendix iv. pp. 698-710, for lists of words peculiar to each New Testament writer.  "The Optative Mode in Hellenistic Greek," by Prof. H. M. Harman, D.D., LL.D. Journal of Soc. of Bibl. Lit. and Exegesis, Dec. 1886, p. 10.  The word hi erourgounta (ha p. leg.) in the highly figurative passage, Romans 15:16, is no exception, being derived not from hi ereus, but from hi eros  But that the style indicates that the Epistle was not actually written by an Alexandrian may be gathered from the non-use of the optative mood. See the reference to Dr. Harman on p. 348.  If the question be asked how Clement of Rome should have been so familiar with this Epistle, the sufficient answer is, that if this Clement be the same with the Clement mentioned in Philip. iv. 3 , as is altogether probable and as is generally asserted in the fathers, they were both companions of St. Paul, though whether they were with him at the same time is not known, and so one of them was likely to know and value the work of the other. Moreover, nearly all the varying traditions about Barnabas concur in speaking of his preaching at Rome, where he would have become personally known to Clement, and whence he may have written this Epistle. If he planted the Church at Milan, as is asserted in the title and proper preface for St. Barnabas Day, in the Ambrosian liturgy, he must have passed through Rome on his way.  Tertullian De Pudicitia, c. 20, Tom. II., fol. 1021, ed. Migne. It is well known that the Pauline authorship of the Epistle was rejected by many of the ancients. Eusebius (Eccl. Hist. vi. 20) mentions that in the list of Caius, Presbyter at Rome (cir. 200), only thirteen epistles of Paul are enumerated, and this is omitted. It is also omitted in the Muratonian fragment, if that be not the same. The Codex Claromontanus (6th cent.) was copied from a ms. not containing Hebrews, but gives at the end of Philemon a stichometrical catalogue of all the books of the Old and New Testaments, and then gives our Epistle. In the catalogue, however, before Revelations and Acts, and immediately after Jude, is mentioned the "Epistle of Barnabas," having 850 lines. It has been conjectured that by this may be meant the Epistle to the Hebrews; for 1 Corinthians put down at 1060 lines, and this would give, in proportion, very nearly the right length for our Epistle, making Hebrews 820 instead of 850 lines, whereas the spurious "Epistle of Barnabas" is nearly one-half longer. (See Salmon, Introd. to the N.T., note at end of xxi., 2d ed., pp. 453, 454.) This conclusion is controverted by Lünemann, Introd. to Heb., sect. i., p. 23, ed. T. & T. Clark.  The following memorandum of the authorities for and against the genuineness of the Epistle of Barnabas, and for its date, has been kindly furnished by the Rev. E. C. Richardson, Librarian of the Hartford Theological Seminary:-- Genuineness, etc . For : Origen, Clement of A., Eusebius, Hieron., Apost. const., Voss, Hammond, Pearson, Bull, Cave, Du Pin, Grynaeus, Wake, Lardner, Fleury, Le Nourry, Russel, Galland, Less, Rosenmüller, Muenscher, Stäudlin, Danz, Bertholdt, Hemsen, Schmidt, Henke, Bleek, Rördam, Gieseler, Näbe, Credner, Bretschneider, Guericke, Francke, Gfrörer, Möhler, Baumgarten-Crusius, De Wette (?), Rysewyk, Schneckenburger, Sprinzl, Alzog, Nirschl, Sharpe. Against : Rivet, Usher, Menard, Daillé, Papebroch, Calmet, Cotelerius, Le Moyne, Tenzel, Natalis Alex., Ittig, Spanheim, Tillemont, Basnage, Oudin, Ceillier, Stolle, Pertsch, Baumgarten, Walch, Mosheim, Semler, Schroekh, Rössler, Starke, Lumper, Michaelis, Gaab, Lange, Hänlein, Winter, Neander, Ullmann, Mynster, Hug, Baur, Winer, Hase, Ebrard, Semisch, Kayser, Reithmayr, Hefele, MacKenzie, Lipsius, Weizäcker, Donaldson, Roberts and D., Riggenbach, Westcott, Braunsberger, Cunningham, Funk, Alford. Interpolated : Schenkel, Heydecke. Date : Reign of Vespasian, Menardus, Ewald, Weizäcker, Milligan; 71-73, Galland; 70-100, Tischendorf (at first); reign of Domitian, Wieseler, Hilgenfeld, Riggenbach. Donaldson, Reuss, Ewald, Dressel, and Ritschl also put it in the first century. Papebroch pronounces for some time later than 97, Hefele for 107-120. Volkmar, Tischendorf (later), Baur, and others, for 119; Tentzel for the reign of Trajan; and Hug, Ullmann, Lücke, Neander, Winer, Zeller, and Köstlin for some time early in the 2d century, while Heydecke distinguishes into a genuine B., 70-71, and an interpolator, 119-121.  The large proportion of quotations from the Psalms in this Epistle is noticed in the article upon it in Smith's Bible Dictionary (where the proportion is stated as 16 out of 32); but my attention was first called to the bearing of this upon the question of authorship by the quick observation of Rev. Hermann Lilienthal. It is not easy to give a precise numerical statement of the proportion because of the large number of historical allusions which can hardly be reckoned as quotations, and also because the New Testament writers often clothe their thoughts in the familiar words of the Old Testament, apparently without any conscious quotation. This matter, however, which cannot be tabulated, is quite in accord with the rest, and the whole Epistle is saturated with the language and the historical allusions of the Psalms. Taking only what may fairly be considered as designed quotations, the relative numbers taken from the Psalms are: Hebrews, 200; St. Luke, 25; St. Paul, 39; all others, 25. The Apocalypse is omitted from the calculation. In the comparatively few quotations in St. Luke less than one-third (17 out of 55) are from the Psalms, and every one of these in recording the words of others; less than one-fifth in St. Paul (16 out of 89); and in the others 22 out of 116. In Hebrews almost exactly one-half.
 Eusebius' Eccl. Hist. vi. 14 (Crusé's translation, p. 213). "But now, as the blessed presbyter used to say, since the Lord who was the Apostle of the Almighty, was sent to the Hebrews, Paul by reason of his inferiority, as if sent to the Gentiles, did not subscribe himself an apostle of the Hebrews; both out of reverence for the Lord, and because he wrote of his abundance to the Hebrews, as a herald and apostle of the Gentiles.'"
 Ibid . The Epistle to the Hebrews he asserts was written by Paul to the Hebrews in the Hebrew tongue, but that it was carefully translated by Luke and published among the Greeks. Therefore one finds the same character of style and of phraseology in the Epistle as in the Acts. "But it is probable that the title, Paul the Apostle, was not prefixed to it. For as he wrote to the Hebrews, who had imbibed prejudices against him and suspected him, he wisely guards against diverting them from the perusal by giving his name."
 Eusebius' Eccl. Hist. vi. 25. Extended quotations from the various writers above referred to, and from many others, may be found in almost any of the innumerable treatises on the subject, and are given with especial fullness and clearness in Alford's Prolegomena.
 Godet in The Expositor, April, 1888, p. 262.
 Bp. George Burgess.
 Journal of the Soc. of Bibl. Literature and Exegesis for June, 1887, pp. 1-27.
 See Thayer's Grimm's N.T. Lexicon, Appendix iv. pp. 698-710, for lists of words peculiar to each New Testament writer.
 "The Optative Mode in Hellenistic Greek," by Prof. H. M. Harman, D.D., LL.D. Journal of Soc. of Bibl. Lit. and Exegesis, Dec. 1886, p. 10.
 The word hi erourgounta (ha p. leg.) in the highly figurative passage, Romans 15:16, is no exception, being derived not from hi ereus, but from hi eros
 But that the style indicates that the Epistle was not actually written by an Alexandrian may be gathered from the non-use of the optative mood. See the reference to Dr. Harman on p. 348.
 If the question be asked how Clement of Rome should have been so familiar with this Epistle, the sufficient answer is, that if this Clement be the same with the Clement mentioned in Philip. iv. 3 , as is altogether probable and as is generally asserted in the fathers, they were both companions of St. Paul, though whether they were with him at the same time is not known, and so one of them was likely to know and value the work of the other. Moreover, nearly all the varying traditions about Barnabas concur in speaking of his preaching at Rome, where he would have become personally known to Clement, and whence he may have written this Epistle. If he planted the Church at Milan, as is asserted in the title and proper preface for St. Barnabas Day, in the Ambrosian liturgy, he must have passed through Rome on his way.
 Tertullian De Pudicitia, c. 20, Tom. II., fol. 1021, ed. Migne. It is well known that the Pauline authorship of the Epistle was rejected by many of the ancients. Eusebius (Eccl. Hist. vi. 20) mentions that in the list of Caius, Presbyter at Rome (cir. 200), only thirteen epistles of Paul are enumerated, and this is omitted. It is also omitted in the Muratonian fragment, if that be not the same. The Codex Claromontanus (6th cent.) was copied from a ms. not containing Hebrews, but gives at the end of Philemon a stichometrical catalogue of all the books of the Old and New Testaments, and then gives our Epistle. In the catalogue, however, before Revelations and Acts, and immediately after Jude, is mentioned the "Epistle of Barnabas," having 850 lines. It has been conjectured that by this may be meant the Epistle to the Hebrews; for 1 Corinthians put down at 1060 lines, and this would give, in proportion, very nearly the right length for our Epistle, making Hebrews 820 instead of 850 lines, whereas the spurious "Epistle of Barnabas" is nearly one-half longer. (See Salmon, Introd. to the N.T., note at end of xxi., 2d ed., pp. 453, 454.) This conclusion is controverted by Lünemann, Introd. to Heb., sect. i., p. 23, ed. T. & T. Clark.
 The following memorandum of the authorities for and against the genuineness of the Epistle of Barnabas, and for its date, has been kindly furnished by the Rev. E. C. Richardson, Librarian of the Hartford Theological Seminary:-- Genuineness, etc . For : Origen, Clement of A., Eusebius, Hieron., Apost. const., Voss, Hammond, Pearson, Bull, Cave, Du Pin, Grynaeus, Wake, Lardner, Fleury, Le Nourry, Russel, Galland, Less, Rosenmüller, Muenscher, Stäudlin, Danz, Bertholdt, Hemsen, Schmidt, Henke, Bleek, Rördam, Gieseler, Näbe, Credner, Bretschneider, Guericke, Francke, Gfrörer, Möhler, Baumgarten-Crusius, De Wette (?), Rysewyk, Schneckenburger, Sprinzl, Alzog, Nirschl, Sharpe. Against : Rivet, Usher, Menard, Daillé, Papebroch, Calmet, Cotelerius, Le Moyne, Tenzel, Natalis Alex., Ittig, Spanheim, Tillemont, Basnage, Oudin, Ceillier, Stolle, Pertsch, Baumgarten, Walch, Mosheim, Semler, Schroekh, Rössler, Starke, Lumper, Michaelis, Gaab, Lange, Hänlein, Winter, Neander, Ullmann, Mynster, Hug, Baur, Winer, Hase, Ebrard, Semisch, Kayser, Reithmayr, Hefele, MacKenzie, Lipsius, Weizäcker, Donaldson, Roberts and D., Riggenbach, Westcott, Braunsberger, Cunningham, Funk, Alford. Interpolated : Schenkel, Heydecke. Date : Reign of Vespasian, Menardus, Ewald, Weizäcker, Milligan; 71-73, Galland; 70-100, Tischendorf (at first); reign of Domitian, Wieseler, Hilgenfeld, Riggenbach. Donaldson, Reuss, Ewald, Dressel, and Ritschl also put it in the first century. Papebroch pronounces for some time later than 97, Hefele for 107-120. Volkmar, Tischendorf (later), Baur, and others, for 119; Tentzel for the reign of Trajan; and Hug, Ullmann, Lücke, Neander, Winer, Zeller, and Köstlin for some time early in the 2d century, while Heydecke distinguishes into a genuine B., 70-71, and an interpolator, 119-121.
 The large proportion of quotations from the Psalms in this Epistle is noticed in the article upon it in Smith's Bible Dictionary (where the proportion is stated as 16 out of 32); but my attention was first called to the bearing of this upon the question of authorship by the quick observation of Rev. Hermann Lilienthal. It is not easy to give a precise numerical statement of the proportion because of the large number of historical allusions which can hardly be reckoned as quotations, and also because the New Testament writers often clothe their thoughts in the familiar words of the Old Testament, apparently without any conscious quotation. This matter, however, which cannot be tabulated, is quite in accord with the rest, and the whole Epistle is saturated with the language and the historical allusions of the Psalms. Taking only what may fairly be considered as designed quotations, the relative numbers taken from the Psalms are: Hebrews, 200; St. Luke, 25; St. Paul, 39; all others, 25. The Apocalypse is omitted from the calculation. In the comparatively few quotations in St. Luke less than one-third (17 out of 55) are from the Psalms, and every one of these in recording the words of others; less than one-fifth in St. Paul (16 out of 89); and in the others 22 out of 116. In Hebrews almost exactly one-half.