ALLEGED OCCURRENCE OF SUPERFETATION. 245 



In all instances in which the discrepancy in size or development between two 

 fetuses is such as to suggest that one of them arose from an ovulation so far removed 

 from the other as to produce unusual differences in size in the later months of 

 prenatal development, the assumption of independent action of the ovaries, con- 

 sidered by King in the case of the rat, would not avoid the foregoing difficulties. 

 Nor could the occurrence of deferred fertilization, which is 'advanced by Sumner 

 as an explanation for the occurrence of supernumerary broods in mice, be used 

 as an explanation in human cases, for the period of vitality of both the ovum and 

 the spermatozoon would seem to be entirely too short to become responsible for 

 marked development differences. 



Sumner found supernumerary litters to occur in 3 per cent of 250 broods in 

 young mice. Such litters were particularly frequent in young deer mice. The 

 frequency of supernumerary broods would seem to vary greatly in closely allied 

 mammals, however, for King observed only 2 cases, and these she regarded only 

 as "seemingly" instances of superfetation, among 700 normal broods of rats. This 

 is an incidence of 0.28, or a frequency only one-eleventh as great as that found by 

 Sumner in mice. Sumner also stated that more than 80 per cent of the super- 

 numerary broods survived nearly or quite nearly to maturity, and King added 

 that all of those broods in rats were raised to maturity and that some were used for 

 breeding purposes later. 



Although both King and Sumner regarded the occurrence of superfetation 

 with considerable reserve, neither seems to have considered the possibility of 

 delayed parturition in one horn of the uterus as worthy of consideration in connec- 

 tion with supernumerary broods between which the time interval was short. 

 Whenever this interval was relatively long as contrasted with the gestation period, 

 such a suggestion is untenable, unless corresponding developmental differences 

 were present. It might be helpful as well as interesting to make a careful gross and 

 microscopic examination of the broods concerned in order to contrast the respec- 

 tive states of development. 



Since L. Fraenkel (1910 b ) found that some rabbits copulate while pregnant, 

 it would be of especial interest to know if supernumerary broods are particularly 

 common in such rabbits, and whether supernumerary broods can be produced 

 in mice should comparable conditions obtain. A particularly interesting question 

 would seem to be raised by the occurrence of such cases as reported by Guerra 

 (1909), in which a second implantation is said to have occurred while a dead 

 fetus still was retained in utero. Under these circumstances the smaller, younger 

 fetus should always be best preserved. Although ovulation, and perhaps also 

 menstruation, could occur under such circumstances because of the occasional regen- 

 eration of the endometrium, even before abortion, one scarcely can regard such a 

 sequence of events as established at present. However, Franco (1910) stated that 

 there is no lack of cases in which a second pregnancy occurred under such circum- 

 stances. But that does not establish the occurrence of superfetation, for in a 

 physiological sense death of a conceptus really terminates a pregnancy. Since at 

 least partial regeneration of the endometrium may occur in the presence of a dead 



