TWINNING IN RUMINANTS 103 



Having shown the untenability of Cole's con- 

 clusions, Lillie presents the results of his own examina- 

 tion of 41 cases of bovine twins, all examined in utcro. 

 The material was collected from the Chica<^() stuekvards 

 by a skilled assistant, and in every case the ovaries of 

 the mother were obtained and the embryonic envelopes 

 of the fetuses were preserved and examined. The 

 nature of the fetal genitaha was determined by dissec- 

 tion, and in some cases by sectioning. No such body 

 of data has previously been obtained on bovine twins. 

 This work is still in progress and will be reported at 

 length in due time. In the meantime it will be well to 

 give here nearly all of the data furnished by Lillie 's 

 abstract in Science. 



Out of 41 cases of bovine twins 14 are both male, 

 21 are of opposite sexes, and 6 are both female. This 

 is about what one would expect, if we interpret the free- 

 martin as a female, for there are about as many same- 

 sexed twins as opposite-sexed. The number of the 

 same-sexed male twins is higher than expected, but 

 perhaps this is due to the comparatively small number 

 of cases. If, according to Hart and Cole, frecmartins 

 also are males, there would be an enormous and inexpli- 

 cable preponderance of males. As Lilhe says: 



The real test of the theory must come from the embr>'ological 

 side. If the sterile freemartin and its bull-mate are monozygotic, 

 they should be included in a single chorion, and there should be 

 a single corpus luteum present. If they are dizygotic, wc might 

 expect two separate chorions and two corpora lutea. 'ihe 

 monochorial condition would not, however, be a conclusive test 

 of monozygotic origin, for two chorions, originally independent, 

 might fuse secondarily. The facts as determined from examina- 

 tion of 41 cases are that about 97.5 per cent of bovine twins are 



