THE TESTICLE AND THE OVARY 369 



theory of the meaning and function of the corpus liiteum is untenable 

 (p. 363). The fact that in a very large number of animals, heat, and 

 presumably, therefore, ovulation, occur at infrequent intervals does 

 not support it, while it has been shown that, in some animals at any 

 rate, ovulation does not take place until oestnis, and consequently 

 that at the time of thp proOestrous hypersemia there are no corpora 

 lutea present in the ovary. These facts, however, are in no way 

 opposed to that part of Fraenkel's theory which assigns to the corpus 

 luteum the function of governing the fixation of the ovum and helping 

 to maintain its nutrition during the first stages of pregnancy. 



Dr. Jolly and the author ^ have carried out a series of experiments 

 upon dogs and rats in which the ovaries were extirpated at different 

 stages during pregnancy, as in Fraenkel's experiments upon rabbits. 

 In the experiments on dogs, ovariotomy was performed at intervals 

 ranging from three days to four weeks after impregnation. The 

 pregnancy was discontinued in every case excepting one, in which 

 a portion of the right ovary which contained the degenerate remains 

 of two undoubted corpora lutea were found post mortem, three days 

 after parturition, when the dog was killed. In this experiment 

 ovariotomy was performed three days after copulation, and parturition 

 occurred fifty days subsequently.' Only a single pup was produced, 

 and birth was premature. The pup died after being suckled normally 

 for three days. The ovaries were also removed from a large number 

 of rats, most of which were in early stages of pregnancy. Pregnancy 

 was continued in no case in which ovariotomy was performed durmg 

 the first six days. In other cases, in which the ovaries were removed 

 at periods varying from the sixth day until near the end of pregnancy, 

 the young were produced normally at full time.^ Control experiments 

 were also earned out in which the abdominal cavity was opened up 

 during an early stage of pregnancy and the ovaries were cauterised, 

 or in which one ovary was removed and not the other, and in these 

 experiments the course of pregnancy was not interfered with.^ We 

 purposely refrained from attempting to extirpate the corpora lutea 

 only while leaving the rest of the ovary, as it appeared to us to be 

 practic3,lly impossible to destroy the whole of the luteal tissue without 

 injuring the entire organ. The ovaries during pregnancy consist 

 very largely of corpora lutea, and any attempt in a relatively small 

 aniinal to discriminate between luteal tissue and stroma, while the 

 ovary was lying in its normal position in the body cavity, seemed in 

 our judgment to be impracticable. 



It will be seen that our experiments on the results of ovariotomy 



' Marshall and Jolly, loc. ait. 



2 In our paper the period of gestation in the rat was wrongly stated to be 

 twenty-eight days. It is in reality about twenty-one days. 

 ' Cf. Carmichael and Marshall, loc. cit. 



