the diversity lies farther back; the manuscript transmission of the title is hopelessly corrupt.

To begin with the codices of the a class, in the Moissac manuscript which, as the other good codices of the same class (with possibly a single exception), contains only the eight books of the Gallic War, this work is ascribed to Suetonius. The text of the Gallic War is preceded by a Latin version of a part of the Antiquities of the Jews by Josephus; the title of the Caesarian work immediately follows the subscription of the other, thus:

OPPINI QVI ET IOSEPHI LIBER X
IVDAICE ANT QVITATIS EXPLICIT

INCIPIT DE TRIMODA OMNIS
GALLIAE DIVISIONE

All this occupies a little less than a quarter of a page of the manuscript, which is written in a characteristic hand assigned by Châtelain (Paléographie des Classiques lat. I, p. 13) to the eleventh century; it fills out the greater part of the space left in the second column (each page being written in two columns) after the last sentence of the text of Josephus, and a new page is begun with the first word. of Caesar's text, Gallia, which is provided with an elaborate initial letter. There is, however, no good reason to doubt that the title is of the same age as the rest of the manuscript.

The Amsterdam codex (Bongarsianus) has a double title: INCIPIT LIBER GAII CESARIS BELLI GALLICI IVLIANI DE NARRATIONE TEMPORVM; then, in red, INCIPIT LIBER SVETONII. The text of this manuscript in general agrees with that of the Moissac codex; the second part of the title, Incipit liber

Suetonii, is probably derived from the common ancestor of the two codices, which is designated by Meusel as x. The parent codex very likely had the fuller form which we find in the Moissac manuscript, which was abbreviated in the Amsterdam codex to make room for the long first part of the title ascribing the work to Caesar, this being borrowed, as we shall see, from a manuscript of the other group of the a class, and inserted as a correction.

The other group of a manuscripts, best represented by Cod. Paris. Lat. 5763 (designated as B by the editors) and Cod. Vatic. 3864, assign the Gallic War to Caesar; the parent codex from which they were derived, designated by Meusel as, evidently had as title INCIPIVNT LIBRI GAII

CAESARIS BELLI GALLICI IVLIANI DE NARRATIONE TEMPORVM.

From this form evidently came the first part of the title in the Amsterdam codex, Incipiunt libri being changed to Incipit liber to accord with the second part of the title, Incipit liber Suetonii, derived from x.

The manuscripts of the B class show less variation in the form of the title; the lost codex to which their origin is traced apparently had INCIPIVNT LIBRI GAII IVLII CAESARIS

BELLI GALLICI DE NARRATIONE TEMPORVM.

With this last form before us it is possible to understand how the awkward Iuliani may have found its way into the title as it appears in the second group of manuscripts in the a class. Without entering into the question of the relative value of the a and ẞ readings in constituting Caesar's text, we may suppose that a scribe or reader of a manuscript in the line of transmission between the archetype and the group had before him, either in his own manuscript or in another to which he had access, the title libri Gaii Iulii Caesaris belli Gallici de narratione temporum; that this seemed to him ambiguous or defective, because it does not necessarily assign the war as well as the literary work to Caesar; that he was familiar with the use of the adjective Iulianus with definite reference to Julius Caesar, as, for example, de Bello Afr. 15 equites Iuliani, ibid. 78 turmas Iulianas, Cic. Phil. xiii, 31 vectigalia Iuliana; that he

therefore corrected the title in his manuscript so as to read libri Gaii Caesaris belli Gallici Iuliani de narratione temporum, intending to convey the meaning, 'the books of Gaius (Julius) Caesar concerned with the recountal of the events of the Gallic War waged by Julius (Caesar)'; and that the manuscript thus corrected transmitted the changed title to its descendants. Such an explanation seems less improbable than that suggested by Hauler (Wien. Stud. XVII, p. 128), which accounts for Iuliani de narratione temporum as originating in a misunderstanding of the title CRONICA IVLII CAESARIS and the opening words of a fragment of the Cosmography of Aethicus Hister thus attributed to Caesar in Cod. Paris. suppl. 685; the same fragment immediately follows the eighth book of the Gallic War in Cod. Vatic. 3864.

As the beginning of the text of the Civil War is lacking in all the manuscripts, we find preceding this work a title of the simplest character, as INCIPIT LIBER PRIMVS BELLI CIVILIS. In the Ashburnham Codex appears DE BELLO CIVILI. INCIPIT LIBER NONVS (Philologus, XLV, p. 214); the first book of the Civil War (including Books i and ii of the current editions), immediately following the eighth of the Gallic War, is here reckoned as the ninth of the Corpus Caesarianum. In a manuscript in the British Museum (Addit. 10084, identified by Holder with Lovaniensis) the subscription of Book viii of the Gallic War and the title of the first book of the Civil War read as follows (cf. Châtelain, op. cit. p. 30): C. Cesaris pontificis maximi ephemeris rerum gestarum belli gallici lib. VIII. expl. feliciter. Iulius Celsus Constantinus v.c. legi tantum. Incipit liber nonus. Little help may be expected from this source for the solution of our problem.

Nor do the subscriptions of the other books yield much of value. The word liber constantly appears, but commentarius, so far as I am aware, only at the end of Book vii of the Gallic War, in certain manuscripts of the a class, as B: Iulius Celsus Constantinus v. c. legi commentarius Caesaris liber septimus explicit. The references to the revision of Julius Celsus Constantinus and of Flavius Licerius Firminus Lupicinus

(at the end of Book ii) raise interesting questions, but contribute no evidence bearing upon the authenticity of any part of the title. A detailed analysis of the subscriptions would be a waste of labor.

Whether the title libri Gaii Iulii Caesaris belli Gallici de narratione temporum descended from the archetype (X) into the a as well as the B manuscripts and in some codex in the line of descent to the x group of the former class was replaced by a title attributing the Gallic War to Suetonius; or whether the first title was confined to the B class and was thence carried over, in a corrected form, to some codex in the line of descent to B and the other manuscripts of the group, there replacing a title, previously common to the a class, in which Suetonius was named as author; or whether, finally, the confusion in the forms of the title as they appear in the manuscripts is to be explained in some other way, it is not necessary, so far as our present problem is concerned, to inquire. It will be sufficient to observe that a title so un-Caesarian in both choice of words and manner of expression cannot possibly have come from the hand of the author. We are therefore justified in adopting another line of approach in order to ascertain, first, whether there is any evidence tending to show that Caesar published his Gallic War anonymously; and in the second place, whether, in case the evidence seems to indicate that it was provided with a title from the beginning, we are able to determine, with any degree of probability, what that title was.

If, as has frequently been assumed, Caesar wrote the Gallic War primarily in order to justify his career of conquest before his fellow countrymen, he might well have thought that something was to be gained by anonymous publication; for if a document containing a favorable view of one side of a controversy can be circulated without a knowledge of its source, it is more apt to be received without prejudice and so to carry greater weight than if it is known to have emanated from a conspicuous partisan. Furthermore, on the supposition that the work was intended to be circulated without the name of the author, we have an adequate explanation of the studied

self-repression of Caesar the writer in always using the third person when referring to Caesar the commander, a circumstance which in later times facilitated the circulation of the work under the name of Suetonius. On the other hand, though by the middle of the first century B.C. the book trade in Rome had begun to be well organized, the references in Cicero's letters are sufficient proof,- if an author not wishing to avail himself of the services of professional copyists and booksellers had prepared a work for private distribution he would, as Dziatzko suggests (Ausgewählte Kapitel des antiken Buchwesens, p. 158), probably have sent the transcripts with a personal note or greeting to the recipients and would not have provided such gift copies with a formal title even though he had had no intention of concealing the authorship. The existence of early manuscripts of the Gallic War without a full title is conceivable, then, upon either of two hypotheses: that of anonymous publication, and that of private distribution; in the latter case, as there was nothing corresponding with our copyright laws, copies might begin to be multiplied and offered for sale as soon as a bookseller should be able to get permission to transcribe one of the gift copies.

We know nothing of the circumstances and manner of composition of the Gallic War except what may be gleaned from internal evidence and from the statement of Hirtius in the preface to Book viii, that Caesar wrote his 'commentaries' with great ease and rapidity. According to the current view the work was composed in the winter of 52-51 B.C., and began to be circulated within a few months thereafter; the place of writing was Bibracte, where, as we learn from the closing chapter of Book vii, Caesar had resolved to spend the winter after the fall of Alesia. At Bibracte, his headquarters, the military records would be available in case he should wish to refresh his memory in regard to details; and though he heard cases there (viii, 4, 2), it might be presumed that he would be better able to command leisure for writing than when in the field or even when sojourning in Cisalpine Gaul. Nevertheless his winter in Bibracte was not unbroken. He could have been at most only a few weeks in camp when