466 
ZOOLOGY: F. R. LILLIE 
The investigation of a remarkable phenomenon in twins of cattle 
furnishes a positive answer to this question so far as the female is con- 
cerned. 
The female of two-sexed twins in cattle, commonly known as the 
free-martin, has long been known to be absolutely sterile as a general 
rule ; however a small percentage of such females is perfectly normal. I 
have found by a study of the embryonic development that the phenome- 
non of sterility is due to fusion of the embryonic membranes of the twins, 
and an anastomosis of the arteries and veins of the female and male 
associates, but more especially of the arteries, s.o that there is literal 
community of blood during foetal life.^ If the anastomosis of the blood- 
vessels does not take place, the female is perfectly normal as is usual 
in the twins or multiple births of all other mammals. In cattle again 
if the twins be of the same sex both are perfectly normal. 
Nature has thus performed here a perfectly controlled experiment, 
which shows that blood community of foetal life between embryos of 
different sex causes steriHty of the female. This fact can be explained 
only on the assumption that the foetal blood carries specific sex-hormones, 
because the only system of the female that is affected is the reproductive 
system. The male, on the other hand, is normal in all its parts, and this 
finds explanation in the fact that the sexual differentiation of the male 
antedates by a little that of the female, and the development of female 
sex-hormones is probably inhibited from the start. 
The time at which the anastomosis of the blood vessels occurs is a 
question of fundamental importance. Sex-differentiation begins in 
cattle when the embryo is about 25 mm. long. The evidence at my 
command indicates strongly that fusion of the embryonic membranes 
begins at about this time or a little earlier; in a pair of twins 15 mm. long 
the membranes overlapped but were not yet fused; in another pair, of 
which one member was 35 mm. and the other 40 mm. in length, the mem- 
branes were perfectly fused so that the place of union could no longer 
be detected. The vascular areas of the two sides overlapped, but the 
larger vessels did not anastomose ; a capillary connection at least between 
the two sides certainly existed. In another pair 50 mm. long the fusion 
was perfect and the anastomosis of the blood-vessels also. Study of 
normal embryos also shows that the conditions precedent to fusion are 
fully established before the 20 mm. stage; and study of the gonads of 
the free-martin shows that the development of the ovarian cortex is 
probably inhibited from the beginning, i.e., from about the 25 mm. stage. 
There can be no doubt that the blood community dates from about the 
time of the sexually indifferent stage, but it certainly varies more or less 
