72 FRANK R. LILLIE. 



is acting against resistance in the female soma ; moreover, this re- 

 sistance is not that of already differentiated parts, for the hormone 

 is introduced before sexual differentiation ; it is rather a constitu- 

 tional resistance native to the determined sex. The phenomena 

 can be understood only on the assumption that the zygotic sex- 

 determining factors are also sex-differentiating factors in mammals 

 as in insects. These factors are reinforced very early by hormone 

 production in the male of mammals, but relatively very late in the 

 female. Such reinforcement can be thought of in the sense of an 

 autocatalytic reaction (as suggested by certain writers, cf. Gold- 

 schmidt, R., Huxley, J., 1920), in which the hormones are regarded 

 as by-products of the sex-determining factors present in the zygote, 

 that accelerate the progress of the differentiation. 



General Discussion. 



On account of certain misunderstandings that appear in the lit- 

 erature since publication of my 191 7 paper, I would like to point 

 out the limitations of the theory of the free-martin. I have so far 

 confined the explanation to the bovine free-martin, though indi- 

 cating the possibility that it may have a wider application. 

 Minoura's experiments (1921) confirm the general applicability of 

 the principle. The explanation is not supposed, however, to apply 

 to all cases of hermaphroditism in cattle; it is not only possible, 

 but probable, that certain cases rest upon a zygotic basis, for 

 hermaphroditism is reported for single births in some cases. The 

 possibility, therefore, exists of an accidental combination of such 

 a case with twinning, though it would necessarily be very rare. 

 There is, however, a sufficient basis for receiving with a certain 

 amount of skepticism cases alleged to be due to twinning in which 

 the history is based on statements of farmers, and not on the per- 

 sonal observations of the investigator. Most of the cases in the 

 literature are, as a matter of fact, of this nature. There is no 

 good reason for rejecting them so long as they are conformable to 

 cases known to be due to twinning; but when they are quite un- 

 conformable, especially if they involve extensive modifications of 

 the external parts, judgment may be reserved until proof is forth- 

 coming that such modifications may be due to twinning. 



I regard the explanation of the bovine free-martin which I have 



