1Diseases of Women I–II (DW) occupy together four hundred pages in the eighth volume of E. Littré’s Oeuvres complètes d’ Hippocrate (1853),1 and despite that editor’s continuous chapter-numbering will be regarded here as independent works. The considerable length of the two treatises, the wide diversity of their contents, and their lack of any easily discernible organizational principle may explain why a need was felt by ancient and mediaeval scholars to help readers navigate and understand the texts. Evidence of this ordering and exegetical tendency is found in all three extant independent Greek manuscripts containing the texts,2

2M provides a complete marginal chapter-numbering system for each of the works – partially aligned to a large collection of majuscule (semi-uncial) subject-headings within its text;

3Θ contains well over 250 majuscule marginalia (glosses and subject headings) distributed over the course of the two works;

4V has acquired a dozen marginal notes in the second half of DW I.

5M

6The chapter-numbering system alluded to is established by an almost complete sequence of Greek majuscule numbers written by the first hand in the margins of each work, β´- ξα´ in DW I and α´- νζ´ in DW II.3 In many chapters, there is no majuscule heading (=title) in the text corresponding to its adjacent marginal number:

DW I: 1-5, 7-14, 16, 31, 37, 39-40 (of 61)
DW II: 1-3, 6-29, 32-33, 35-39, 41, 54-55 (of 57).

7Other chapters, however, have majuscule headings at the beginning of their minuscule text, often introduced with the word περὶ. These M chapter-headings can be divided into two groups according to their manuscript transmission:

8For chapters DW I 38, 41-42 and 44-61 and for all of DW II the headings are present not only in M, but also (independent of any numbering system) within the texts of the other two independent manuscripts, Θ (usually written in minuscule letters, but occasionally in semi-uncial) and V (always in minuscule). Besides these numbered chapter-headings, M also contains over one hundred unnumbered majuscule headings in its text between 156,3 L and 404,7 L (with a gap between 266,10 L and 352,15 L) that have more or less fully verbatim correspondences within the minuscule text of Θ and V.

9For chapters DW I 6, 15, 17-30, 32-36 and 43 the headings are found only in the manuscript M: these chapter-headings unique to the M manuscript tradition I will designate M^add on the supposition that they were added to the M tradition after its separation from the common fifth or sixth century MΘV archetype4 rather than lost from both the Θ and V traditions.

10These are:

11DW I:

6 (30,8 L) ϛ´ περὶ μέτρου κενώσεως ἐμμήνων καὶ περὶ τῶν ὑγιαινουσῶν
15 (82,9 L) ιε´ περὶ ἄφθας παιδίου
17 (96,6 L) ιζ´ περὶ ἑλκώσεως καὶ παρατρίψεως αἰδοίου γυναικὸς ἐν τόκῳ
18 (100,15) ιη´ περὶ ῥοῦ ἐκ τόκου
19 (100,20 [marg. sup.] L) ιθ´ περὶ αἵματος ἀναγωγῆς ἐκ τόκου
20 (102,6 L) κ´ γάλακτος γεννητικά
21 (104,3 L) κα´ περὶ τῶν τὴν λοχίαν κάθαρσιν προτρεπόντων
22 (104,10 L) κβ´ περὶ θρομβωθέντων λοχίων
23 (104,15 L) κγ´ περὶ ὑστέρου δυσαπολύτου
24 (106,5 L) κδ´ περὶ χορείου μὴ συνεκκριθέντος τῷ βρέφει
25 (106,14 L) κε´ περὶ ἀμβλώσεως δυσεκκρίτου
26 (106,18 L) κϛ´ περὶ χορείου ἐναπολειφθέντι ἐν ἀμβλωσμῷ
27 (108,3 L) κζ´ περὶ ἑλκώσεως μήτρας καὶ στομίου φλεγμήναντος ἐκ τόκου
28 (108,8 L) κη´ περὶ φλεγμονῆς ὑστέρων ἐκ τόκου
29 (112,6 L) κθ´ περὶ φλεγμονῆς
30 (112,14 L) λ´ περὶ φλεγμονῆς μετὰ ἐπισχέσεως λοχείων
32 (116,19 L) λβ´ περὶ ὑδέρου ἐν μήτρᾳ
33 (118,18 L) λγ´ περὶ ὑδέρου ἐν κυήσει
34 (122,6 L) λδ´ ὅπως ὁ σπλὴν ὀγκοῦται κατὰ τοὺς πυρετοὺς ὑπὸ τῶν ποτῶν
35 (124,15 L) λε´ περὶ ὄγκων δι’ ἐπίσχεσιν
36 (126,20 L) λϛ´ περὶ ἑλκώσεως μήτρας
43 (170,9 L) μγ´ περὶ δυστοκουσῶν.

12Besides these headings added in M in relationship to its numbering system, there are eight further, otherwise unknown majuscule additions (also M^add) in M’s text or occasionally margins unaccompanied by chapter numbers:

44,24 L ὅτι δεῖ πρὸ τῶν καθαρσίων τὰς μήτρας προπυριᾶν
64,5 [marg. inf.] L περὶ τῶν ἀπορρεουσῶν τὴν γονήν
66,15 [marg. sup.] L τίνων ἀπέχεσθαι δεῖ τὴν κύουσαν
88,17 L πόσα συμπάσχει τῇ μήτρᾳ
146,19 L περὶ ἐμβρυουλκίας τῶν τεθνεώτων
152,22 [marg. sup.] L γάλακτος γεννητικά
196,17 L περὶ φλεγμονῆς ἐκ τόκου
212,13 L καθαρτικὸν ἄτοκον καθῆραι ἢν τὸ στόμα ὀρθῶς ἔχῃ.
It is to the unique M^add material that I will return below.

13Θ

14In the minuscule manuscript Θ the first hand has, apparently at the time of its original creation, included an extensive series of majuscule (semi-uncial) marginal notes (=Θ^marg) to the texts of DW I-II. These side-notes, which are often introduced by the letters ΣΗ signifying σημείωσαι = nota bene, can be divided according to their purpose into three groups: a) glosses, b) references and c) headings.

15Glosses: between 20,17 L and 116,19 L seventeen original Θ^marg glosses on what were apparently judged to be difficult expressions in the text are routinely introduced with the word ἀντὶ:5

1. 20,17 L ἀντὶ ἀπολαύει καὶ ἀναπίνει (alternatives for ἐπαυρίσκεται)
2. 32,4 L ἀντὶ τοῦ κενὸν ἐξ ἀσιτίας γένηται καὶ τὸ πᾶν μὲν σῶμα, μάλιστα δὲ γαστὴρ καὶ ἔντερον (an explanation of κενεαγγήσῃ)
3. 32,11 L ἀντὶ τοῦ τὴν ἀνάπνοον· κοιλίαν δὲ τὸν θώρακα λέγει (an explanation of διάπνοον τὸν περὶ τὴν κοιλίην)
4. 32,13 L ἀντὶ εἰς τὰ ἔντερα (a clarification of ἐς τὰ ὑποχόνδρια)
5. 32,15 L ἀντὶ τοῦ εἰς τὴν οἰκείαν χώραν (a clarification of ἐς χώρην)
6. 32,16 L ἀντὶ τοῦ ψόφος τις (a gloss on τρυσμός)
7. 34,21 L ἢ ἐξ ἀναχύσεως (an explanation of τότε δὲ πνὶξ προσπεσεῖται ἰσχυρή)
8. 36,6 L ἀντὶ τοῦ οὐ δύναται (a clarification of the import here of οὐκ ἐθέλει)
9. 36,14 ῥόον ἐνταῦθα τὸ πάθος νοῶν (a clarification of the significance of ῥόον here)
10. 40,3 ἀντὶ τοῦ κριθῶν πλύματος (a clarification of the meaning of κριθέων here)
11. 42,19 L ἀντὶ τοῦ ἀκριβεῖ (a gloss on σκεθρῇ)
12. 48,7 L ἀντὶ τοῦ τῶν ἀγριμίων (a gloss on θηρίων)
13. 66,11 L ἀντὶ τοῦ σφυγμοί (a gloss on the word παλμοί)
14. 80,5 L ἀντὶ τοῦ ἐκδυναμωθῇ (a gloss on ἐξανεμωθῇ)
15. 110,7 L ἀντὶ τοῦ ἀμυγδάλινον (a gloss on νέτωπον)
16. 112,6 ἀντὶ τοῦ ἐν τόκῳ βιαίῳ (a clarification of πίμπραται ἡ κοιλίη καὶ μεγάλη γίνεται)
17. 116,19 ἀντὶ λανάτα (a gloss on ῥήνικας).

16E. Nachmanson’s assessment of these glosses will be presented in more detail below.6

17References: a few of the notes are references that draw attention to important points made in the Hippocratic text; these are introduced by the word ὅτι:

128,5 L ΣΗ ὅτι ἐκ τρωσμοῦ ἡ νόσος
150,9 L ΣΗ ὅτι τὸ μὲν ἄρρεν γ´ μηνῶν, τὸ δὲ θῆλυ δ´ μηνῶν κινεῖται.

18Headings: by far the most frequent marginalia, however, are over two hundred subject-headings that extend regularly from 20,4 L to 404,7 L, with a gap between 280,12 L and 306,3 L: e.g.

176,10 L ΣΗ μητρῶν καθαρτικῶν· ὅταν παιδίου ἐναποθάνοντος, αἷμα ἐμμείνῃ
182,13 L ΣΗ χορίον ἐξάγει καὶ ἐπιμήνια κατασπᾷ καὶ ἔμβρυον ἡμίεργον ἐξάγει
216,11 L ΣΗ κλυσμὸς ἑλκώσεως μητρῶν μετὰ στραγγουρίας
394,5-17 L πρόσθετα ὕδωρ ἄγοντα· καὶ μύξας καὶ δέρματα καὶ οὐχ ἑλκοῖ· καὶ ἰχῶρας ὑφαίμους καὶ αἵματα.

19These majuscule subject-headings all derive from the Hippocratic text to which they are referring, but display varying degrees of proximity, some being simply repetitions or summaries of the text that retain the wording of their model (cf. the four examples between 176,10 L and 394,5-17 L just presented), others are short newly composed subject-headings: e.g.

316,1 L, etc. ἄλλη (νόσος) καὶ θεραπεία (over 25 times)
174,16 L, etc. ἄλλο (over 10 times)
218,4 L, etc. ἔτι περὶ τοῦ αὐτοῦ (5 times)

20or the series

258,4 L περὶ ῥόου λευκοῦ
260,23 L περὶ τοῦ αὐτοῦ ῥόου
262,12 L and 264,21 L περὶ ἄλλου ῥόου,

21and yet others are freer paraphrases of the text employing their own vocabulary: e.g.

32,1 L ΣΗ περὶ πνιγὸς τῆς ἐξαναδρομῆς αὐτῆς τῆς ὑστέρας
54,17 L παλίν περὶ τῶν ἐξυγρασμένων
62,4 L ΣΗ κοινὰς τινὰς πάλιν διδασκαλίας περὶ τῶν ἀφόρων
66,8 L ΣΗ περὶ τῶν ἀπὸ κεφαλῆς ῥευματιζομένων ἐγκύων. καὶ τῆς ἐντεῦθεν ἀμβλώσεως
76, L περὶ πνιγὸς ἐπὶ ἐγκύου αἰφνιδίου αἰτίων καὶ συμπτωμάτων
78,1 L περὶ κοινῶν συμπτωμάτων τῶν ἐγκύων καὶ τῶν αἰτίων αὐτῶν
94,11 L περὶ τῆς διὰ φλεγμονῆς τῆς μήτρας ἐπισχέσεως τῶν λοχειῶν
110,11 L ΣΗ περὶ τῶν αὐτῶν τῆς μήτρας πόνων καὶ φλεγμονῆς μεγάλης τοῦ ὑπογαστρίου συμπτώματός τε καὶ θεραπείας
148,24 L ΣΗ περὶ μύλης κυήσεως αἰτίων τε καὶ συμπτωμάτων· ἤ τε καὶ σημειώσεως καὶ θεραπείας
234,1 L ΣΗ περὶ ῥόου λευκοῦ καὶ πυρροῦ καὶ ἐρυθροῦ· ἐν ποίαις ἑκάσταις ἡλικίαις μάλιστα γίνεται
308,14 L ΣΗ περὶ διαφόρων νοσημάτων τῶν ὑστέρων. ἐκ τοῦ ἰδίου τόπου μετακινουμένων συμπτωμάτων τε αὐτῶν καὶ θεραπείας
358,19 L ΣΗ περὶ ἐμπνευματώσεως μητρῶν καὶ θεραπείας.
It is to the original textual material of Θ^marg―present in both the glosses and headings―that I will return below, that is, to texts present nowhere else in the Hippocratic manuscript tradition.

22V

23The manuscript V contains the following marginal notes (=V^marg) in various hands between 136,1 L and 216,11 L, that is, in the section of DW I copied in Vb:7

136,1 L κοινοὶ λόγοι περὶ ἑλκῶν (cf. Θ^marg 62,4 and 78,11)
140,14 L περὶ ἑλκῶν
146,19 L περὶ κακοσχήμων καὶ τεθνεώτων (cf. M^add)
148,24 L περὶ μύλ<ης>
152,22 L γάλακτος γεννητικά (cf. M^add)
170,9 L περὶ Δυστοκουσῶν (cf. M^add)
178,1 L πρόσθετ. (sic)
196,17 L περὶ φλεγμονῆς ἐκ τόκου (cf. M^add)
198,21 L περὶ κλυσμῶν
202,22 L περὶ κλυσμῶν
214,19 L Γάλα
216,11 L κλυσμός.

24Four of these V^marg readings have parallels among M^add readings as noted above: at 152,22 L, 170,9 L and 196,17 L the match is perfect; at 146,19 L the readings are close enough to suggest a common source, but far enough to exclude a direct dependence:

M^add περὶ ἐμβρυουλκίας τῶν τεθνεώτων
V^marg περὶ κακοσχήμων καὶ τεθνεώτων.

25One other V^marg marginal note displays a certain proximity of expression with two Θ^marg readings:

V^marg 136,1 L κοινοὶ λόγοι περὶ ἑλκῶν
Θ^marg 62,4 L κοινὰς τινὰς πάλιν διδασκαλίας περὶ τῶν ἀφόρων
Θ^marg 78,11 L περὶ κοινῶν συμπτωμάτων τῶν ἐγκύων καὶ τῶν αἰτίων αὐτῶν.

Interpretation of M^add, Θ^marg and V^marg

26The Corpus Hippocraticum (CH) provided an important point of reference for subsequent Greek medical theory and practice from at the latest the time of Diocles of Carystus and Herophilus, both directly and through the lenses of its many interpreters, such as the empiric Apollonius of Citium, the dogmatist Rufus of Ephesus, the methodist Soranus of Ephesus, and the eclectic Galen of Pergamum; in the sixth century of our era Stephanus of Athens wrote Greek commentaries on the Aphorisms and Prognostic.

27I will now examine the new material provided by M^add, Θ^marg and V^marg with regard to its origins and its significance for our understanding of how the DW I-II texts were handled in the period between antiquity and their transcription to the extant independent manuscripts.

28Language usages not found in the Corpus Hippocraticum (CH):

29According to Liddell and Scott 1940 (=LSJ), the earliest occurrence of the word σύμπτωμα in the sense “symptom” (as defined in Murray 1933),8 is in Philodemus’ De Ira, which dates from the first century B.C. In the CH symptoms are regularly called σημεῖα, occasionally τεκμήρια, or often simply ταῦτα or τάδε in reference to preceding or following lists. In his headings referring to the contents of specific texts, Θ^marg refers to collections of symptoms on seventeen occasions as συμπτώματα.9

30Pregnancy and the pregnant woman are routinely referred to in the CH by variations of the expressions ἐν γαστρὶ λαμβάνειν and ἐν γαστρὶ ἔχειν. Soranus and Galen use this language only in direct quotations from the CH, whereas in their own texts they employ forms of the verb κύειν such as ἡ κύουσα.10 The adjective ἔγκυος (“pregnant”) which is used by Herodotus, occurs once in the CH, never to my knowledge in Soranus or Galen, but several times in the Aphorisms commentary of Stephanus.11 M^add and Θ^marg never take over Hippocratic ἐν γαστρὶ expressions from the texts they are referring to, but in their place M^add employs ἡ κύουσα once12 and Θ^marg makes use of ἡ ἔγκυος seven times.13

31The word διάθεσις occurs eleven times in the CH with the meaning well exemplified by a passage in Ancient Medicine 7: “the condition, in which an individual happens to be on any occasion” (ἡ διάθεσις, ἐν οἵῃ ἂν ἑκάστοτε ἕκαστος τύχῃ διακείμενος). Such conditions include the “moist” (Diseases III 15), the “athletic” (Nutriment 34), the “healthy” (Precepts 5), and the “painful” (Precepts 7), but never actual diseases, for which the authors of the Hippocratic writings routinely employ the terms νοῦσος, νόσημα, and πάθημα, and occasionally πάθος. Θ^marg, however, employs the term διάθεσις twice as an exact parallel to εἶδος τοῦ νοσήματος in referring to two of four different abnormal menstrual fluxes being described in the text―one through the vagina, one flowing out above the groin, one forming a growth in a flank, and a fourth expelling pus and menses through the flank―clearly referring not to temporary states, but to actual forms of disease:

20,4 L ΣΗ περὶ ἄλλης διαθέσεως
20,11 L ΣΗ περὶ ἑτέρας διαθέσεως
20,22 L ΣΗ ἄλλο πάλιν εἶδος τοῦ νοσήματος.

32This migration of the term διάθεσις from the realm of pathogenesis, where it plays a key role in the medical writings of Galen and his successors, to one more like morbid anatomy, reflects a general development of medical thought seen in such writers as Stephanus of Athens in the sixth century, where for example the traditional clinical definition of consumption is replaced by an anatomical one: e.g. “What consumption is, we have learned on several occasions: it is ulceration of the lungs.”14

33The word κενεαγγήσῃ (“vessel[s] are/become empty”), which is glossed by Θ^marg as recorded under the number 2 in the list above, is followed a few lines later in the Hippocratic text by a clarification of which vessel is meant: “the cavity (κοιλίη) is empty”. Since the common Hippocratic term for the gastro-intestinal tract, κοιλίη, was long obsolete by the time of Θ^marg, the scholiast has here clarified it for his readers by adding the explanation: “in particular the stomach and intestine”.

34Words that do not occur in the CH:

35In the following chart all the words present in M^add, Θ^marg and V^marg that do not occur in the Hippocratic writings are recorded in alphabetical order; after each word the name of the author to whom LSJ attributes the earliest use is recorded in square brackets. The next three columns give Littré (volume 8) page and line references to occurrences in each of the three sources; finally, in the last two columns the presence of the words in the Hippocratic commentaries of Galen and Stephanus is indicated with an “x”.

36The examples in this chart illustrate the evolution of medical terminology from the time of Xenophon (ca. 430-ca. 355) to that of Athenaeus of Naucratis in the second century A.D, or the Hippiatrica even later. The frequency with which these new words occur in the Hippocratic commentaries of Galen and Stephanus testifies to their established place in the medical vocabulary of the second and sixth centuries respectively.

37Words not included in LSJ:

38The following seven words occur in Θ^marg at the places noted, but are not recorded in LSJ:

ἀγλαϊστικός 360,18 L           ἐξανάχυσις 34,21 L

ἀγρίμιος 48,7 L15                  λανάτον 116,19 L16

ἐκδυναμόω 80,5 L17              μυλίασις 360,18 L

ἐξαναδρομή 32,1 L

39E. Nachmanson, in his assessment of the significance of the Θ^marg glosses for the study of Erotian’s Glossary gives the following account:18

Einige kurze Worte genügen zur Charakterisierung dieses Scholiasten. Zunächst N. 14 ἐκδυναμωθῇ: das leicht verständliche Wort, das, soviel ich sehe, in den Lexica unbelegt ist, scheint ad hoc fabriziert zu sein. N. 17 λανάτα: … ist nach Du Cange, Gloss. med. et inf. Graec. s.v. (der insgesamt drei Belege gibt) zuerst aus dem Archimandrit Dorotheos in Palästina (ca. 700 nach Krumbacher, Byz. Litt.-Gesch. S. 145 f.) belegt. Auch andere triviale Erklärungen dürfen billig dem Scholiasten selbst zugetraut werden, wie beispielsweise N. 5 und N. 12. Jedenfalls ist es ziemlich aussichtslos, nach bestimmten Quellen zu suchen, wenn sich auch die Erklärungen gelegentlich mit sonst vorhandenen berühren; man könnte ja zu N. 1 Hesych II 137,76, zu N. 8 Galen XVIII A, S. 651,8 K., zu N. 11 Erotian 119,11 oder Galen XVIII A, S. 573,11 K. oder Hesych IV 40,83 vergleichen. Zwar einige Brocken von eingehenderer Gelehrsamkeit hat der Scholiast – aus irgend einem bequem zugänglichen Handbuch? – aufgeschnappt …
Was nun unsere Zwecke angeht, so ist der Scholiast, wie ich schon oben sagte, offenbar kein Gewährsmann für Erotianea und, wie bereits Ilberg, Abh., S. 124 kurz hervorhob, war Klein völlig im Unrecht, als er zwei dieser Erklärungen (N. 3 und 4) den Erotianfragmenten (als N. LVI, LVII) einverleibte.

40Nachmanson is certainly correct in concluding that Θ^marg has nothing to contribute to an Erotian edition, but just what disqualifies this source for Erotian scholarship qualifies it for the study of Hippocratic scholarship between antiquity and the time of Θ. When Nachmanson tells us Θ^marg is creating words ad hoc, he is recognizing an original scholarly contribution.

41Although the time and circumstances of the origin of the M^add and Θ^marg readings cannot be ascertained with any precision, the following considerations may be useful in delimiting the range of possibilities to a certain degree.

42The fact that both these sources are written in semi-uncial letters suggests that they were transcribed directly from models where both they and the Hippocratic texts they are illustrating were in majuscules, i.e. manuscripts from before the transliteration that took place beginning in the eighth/ninth century.

43The few missing marginal numbers from M’s numbering system (DW I, 1 and 59; DW II, 4, 11, 12, 13, 14, 15 and 31) provide evidence that this numbering system was not devised by the M scribe himself, but by one of his predecessors.

44The uniquely transmitted majuscule headings in M^add and Θ^marg, although created for the same purpose, have virtually no verbal correspondences, the one exception being easily explainable as the coincidental choice of an obvious circumlocution.19 This situation confirms the complete independence of the parallel M^add and Θ^marg exegetical projects.

45M, or more likely one of its ancestors has left a slight trace of its additions in a few marginalia of V. Since these scholia are not in the hand of V, it seems likely that this importation occurred in V itself.

46Θ’s marginal notes seem to be without any echo at all, which is not surprising since Θ itself has left little or no other trace in the later manuscript tradition.

47The collection of thirty non-Hippocratic words found in M^add and Θ^marg (two of which are present in both sources, but not in the same passage, and many of which are part of the standard medical vocabulary of Galen and Stephanus)20 serve only very generally to locate the time of their composition between the first and eighth centuries, but whereas M^add employs no word unrecorded in LSJ, Θ^marg includes seven such (two of which have citations in the LBG ascribed to the seventh and eleventh centuries respectively, and one of which entered the modern Greek language), suggesting that its scholiast was active later than the author of the M^add headings.

48Practices of the scholiasts from whom the M^add and Θ^marg additions originated:

49As the material presented above demonstrates, both of these sources employed an often shared set of linguistic practices which may be briefly summarized as follows:

• a) The provision of glosses and explanations: for M^add e.g. περὶ ἑλκώσεως καὶ παρατρίψεως αἰδοίου γυναικὸς for ἢν δ’ ἐκ τόκου ἐούσῃ συμφραχθῇ τι τοῦ αἰδοίου; περὶ αἵματος ἀναγωγῆς ἐκ τόκου for ἢν δὲ αἷμα ἐκ τόκου ἐμέῃ; for Θ^marg see the glosses listed above.21
• b) The replacement of Hippocratic vocabulary by later expressions: e.g ἄμβλωσις or ἀμβλωσμός for τρωσμός; ἡ ἔγκυος or ἡ κύουσα for ἡ ἐν γαστρὶ ἔχουσα; σύμπτωμα for σημεῖον.22
• c) The replacement of common Hippocratic expressions with more abstract or technical ones: e.g. ἀπόσβεσις derived from ἀποσβέννυμι; δυστοκία derived from δυστοκέω; ῥευματίζομαι derived from ῥεῦμα; σημείωσις derived from σημεῖον.23
• d) The employment of new, technical compound words: e.g δυσαπόλυτος [Erotian], δυσέκκριτος [Athenaeus], ἐμβρυουλκία [Archigenes], ἐμπνευμάτωσις [Soranus], κακόσχημος [Herodian Gr.].24
• e) Reformulation of the Hippocratic text: e.g.
  M^add 66,15 L τίνων ἀπέχεσθαι δεῖ τὴν κύουσαν for εἰσι κίνδυνοι οἵοις τὰ ἔμβρυα φθείρονται.
• f) M^add 152,22 L γάλακτος γεννητικά for τὸ δὲ γάλα ὅκως γίνεται.
• g) Θ^marg 32,1 L περὶ πνιγὸς τῆς ἐξαναδρομῆς αὐτῆς τῆς ὑστέρας for ἢν δὲ πνὶξ προστῇ … αἱ μῆτραι … στρέφονται.

Conclusion

50I hope to have shown that, at some unknown time between perhaps Galen and the copying of the majuscule manuscripts from which M and Θ were transcribed, the Hippocratic treatises DW I-II attracted the attention of certain unknown scholars who each developed a series of remarks designed to make the texts more accessible to readers; neither of the two series of new material gives any evidence of an awareness of the existence of the other.

51Of M’s thirty majuscule textual additions, most were carried along―albeit in minuscule―into the twelfth century manuscript Parisinus Graecus 2140 (I), some persisted through the recentiores and the early printed editions of the text, but only three arrived in the edition of Littré, two in the text:

196,17 L περὶ φλεγμονῆς ἐκ τόκου
212,13 L καθαρτικὸν ἄτοκον καθῆραι ἢν τὸ στόμα ὀρθῶς ἔχῃ,

52and one in the critical apparatus:

170,9 L περὶ δυστοκουσῶν.

53Of the Θ marginalia, only the glosses attracted later scholars’ notice, fourteen of them being recorded in Littré’s apparatus.25

54In sum, the evidence presented in the present paper is meant to suggest that these neglected sources deserve more attention than they have hitherto received, as primary witnesses of late ancient and early byzantine medical scholarship.