In twins from two distinct ova there is no difficulty in seeing that the souls are placed in these in the same manner as the soul is put in the normal single embryo. When the twins come from one ovum the condition is not so simple. The oval nucleus is the essential part that goes from the maternal side, and human ova at times contain two nuclei, as occasionally hens' eggs do; a double-yoked hen's egg has two nuclei, and two nuclei have been found in a single yolk. Kölliker, Stöckel, and von Franque have observed double germinal vesicles in single human ova. In such a condition two spermatozoa could fecundate the two nuclei and the development go on as in the case of twins from distinct ova. There is a theory which holds that homologous twins (uni-oval) can develop from a single germinal vesicle which splits into two primitive streaks and two gastrulas. According to this opinion, if the germinal vesicle divide entirely, two fetuses develop which are always of the same sex, and which resemble each other so closely in appearance that it is very difficult to differentiate them. This theory holds also that should the germinal vesicle not split fully, the lack of fission causes the various kinds of double monsters. The germinal vesicle that supposedly splits into two is not fecundated by two spermatozoa, they say, because where there is only one nucleus in the beginning, the entrance of a second spermatozoön commonly kills the ovum. This last assertion has been disproved of late. Some followers of the splitting theory hold that double monsters arise from the union of two originally separate primitive traces (_Verwachsungstheorie_). Others say that a single primitive trace of blastoderm cleaves more or less thoroughly and makes the double monster (_Spaltungstheorie_). The earliest human double monster (Ahlfeld's case) was in the fourth week of gestation; therefore whatever is held in these theories as regards human monsters is only through analogy with lower animals. Gerlach[57] saw bifurcation at the cephalic end of a chicken embryo sixteen hours old. In this case the first change was a broadening of the anterior end of the primitive streak; next a forked divergence appeared, and by the twenty-sixth hour the bifurcation was half as long as the undivided posterior part. Whether this was a case of two nuclei or not is not known. [57] _U. d. Entstchungsweise der vordern Verdoppelung. Deutsch. Archiv. f. klin. Med._, 1887.