380 



of forming two embryos. Perhaps in cases of identical triplets in 

 man, the three embryos arise through a division of one of primary 

 buds, while the other bud remains undivided: and likewise in 

 identical quadruplets, both primary buds would divide, as it is the 

 case in the Texas armadillo. Undoubtedly the few five-embryo sets 

 that have been found in our species of the armadillo have come about 

 through the production of three embryos from one of the primary buds, 

 and two from the other bud; for in all such cases the embryos are 

 arranged into two groups, of two and three embryos respectively. 

 Finally, Fernandez^) has shown that the South American armadillo 

 (Tatu hybridum) produces a variable number of embryos at a litter, from 

 7 to 12, and a study of his Figs. 1 and 2 indicates that the conver- 

 ging amniotic tubes leading from the several embryos can be arranged 

 into two groups, each of which joins the common amniotic vesicle 

 by practically a single channel. This would at least suggest that the 

 several embryos arise from two primary buds, each of which forms 

 from three to six secondary buds. 



The argument in favor of regarding polyembryony as the result 

 of a precocious budding rather than as a result of the separation of 

 blastomeres is not based on a study of this phenomena in mammals 

 alone ; for in the most striking cases reported in the literature, poly- 

 embryony woald seem to be connected in some way with a budding 

 process. Thus the work of both Marchal and Silvestri on the 

 parasitic hymenoptera indicates that the actual production of the 

 embryos can be interpreted as a type of budding; and again, in the 

 Bryozoa, Harmer and Robertson point out that budding takes place. 

 Finally, in the well known case of twin formation in the earthworm, 

 Kleinenberg has shown that the gemelliparous condition there is 

 brought about through a sort of budding, or division of the gastrula. 



It is a striking fact that in all of these well known cases of 

 polyembryony the early cleavage of the egg is indeterminate, so that 

 it would not be possible for one to trace out a cell-lineage for any 

 particular embryo. 



These facts bring us to a brief consideration of the causes under- 

 lying polyembryony. In all of the cases mentioned above the embryos 

 lead a parasitic mode of life, or at least are placed in an environment 

 which supplies an abundance of foodstuffs. It is difficult to avoid 



') Fernandez, M., Morphologisches Jahrbuch, Bd. 39, Heft 2. 



