52 FRANK R. LILLIE. 



to demonstrate these points. This may be a case of monozygotic 

 twinning; but the possibility that the Graafian follicle furnishing 

 the corpus luteum contained two ova, both of which were fertilized 

 and developed, can not be neglected. 



My records include 36 cases of twins in which both maternal 

 ovaries were present; of these 35 had two corpora lutea either in 

 both ovaries or in one. In the 90 cases reported by Keller and 

 Tandler there were 88 cases of two corpora lutea and 2 of three 

 corpora lutea. Thus only one case of 126 known cases exhibits 

 one corpus luteum for two embryos. These statistics demonstrate 

 how futile it is to continue to appeal to monozygotic twinning, as 

 D. Berry Hart (1918) does, as the cause of the intersexual con- 

 dition of the free-martin. 



4. The Formation of the Free-martin an " All-or-none " 



Reaction. 



Cases of two corpora lutea in one maternal ovary are apparently 

 not very rare in cattle twins, to judge from the statistics of Keller 

 and Tandler, who found 36 such cases out of 90. I have found 

 only three such cases out of 36 cases where both ovaries were 

 present, but the small number may be due to the fact that the 

 butchers, on whom we relied for selecting uteri containing twins, 

 would be likely to overlook twin pregnancies confined to one horn 

 of the uterus. Cases of this kind in which one foetus is a male 

 and the other a free-martin are of considerable theoretical interest, 

 because implantation of the twins in one horn of the uterus offers 

 presumably the maximum opportunity for modification of the 

 female. 



In all cases in which the twins come from two ovaries it is 

 necessary for the chorionic vesicles to grow entirely through their 

 respective horns of the uterus before they can meet in the corpus 

 uteri, and the fusion that takes place there must occur near the 

 apices of the vesicles, as far removed from the embryos as possible. 

 It is probable that the embryos would be about 15 mm. long before 

 fusion (cf. my earlier paper, p. 389), and we have no data for 

 deciding how quickly vascular anastomosis would become estab- 

 lished and effective. When, on the other hand, both ova come 

 from one ovary (Fig. 1) and develop in one horn of the uterus, 



