NOVEMBEB 16, 1917] 



SCIENCE 



487 



June, 1913 (Biol. Bull.), in whicli he records 

 his observations on certain non-fertilized eggs, 

 in which cleavage advanced parthenogenetically 

 as far as the eight-cell stage, apparently in nor- 

 mal manner. In this chapter the gastrulation, 

 the germ-layer inversion, and the formation 

 of first two and then four embryos from the 

 embryonic area, are given in order, followed 

 by the subsequent separation of the four dis- 

 tinct embryos, each with its own amnion and 

 placenta. Corresponding to their origin, two 

 secondary embryos from the two primai-y ones, 

 the four are distinctly paired, the two of each 

 pair revealing a more complete identity than 

 does either one when compared with a mem- 

 ber of the other pair, and this relationship in 

 certain extra-embryonal features, such as the 

 approximation of the placentas of each pair, is 

 shown in anatomical relations up to birth. 



The condition in other species and genera 

 of armadillo is presented in Chapter III., 

 which shows that the number of young varies 

 from eight (occasionally 7-12), polyembryonic 

 ones in Dasypus hyhridus, to Euphractus vil- 

 losus, which is not polyembryonic, but produces 

 fraternal twins from two separate eggs, or, oc- 

 casionally, bears only a single young. The facts 

 for this chapter are furnished largely by the 

 work of Fernandez of the Museo Nacional at 

 La Plata (Argentina), who has made special 

 studies on the armadillos of South America, 

 and whose first account of the polyembryony 

 of Daypus hyhridus appeared in 1900 (Morph. 

 Jahrh.) almost simultaneously with the first 

 paper of Newman and Patterson on the same 

 phenomenon in D. novemcinctus (Biol. Bull.). 



Chapter TV., although short, has a special 

 interest since in it the author discusses causes 

 of polyembryonic development, thereby bring- 

 ing in something of the many theories that 

 have been brought forward to account for hu- 

 man twins, at least those of the duplicate or 

 monochorial type. The author considers the 

 phenomenon one of fission, " if by fission we 

 mean merely the physiological isolation of sev- 

 eral secondary points in a single embi-yonic 

 vesicle, and the consequent acquisition by 

 these points of independence in growth and 

 development" (p. 93). He assumes a consid- 



erable amount of differentiation to have oc- 

 curred before these points become isolated, " so 

 that genetic factors are unequally distributed 

 in the various regions which give rise to the 

 new apical points," and thus if two embryos 

 are developed from closely adjacent territory 

 they are likely to be more nearly alike than 

 those which are a greater distance apart on 

 the blastoderm. This accounts for the phe- 

 nomenon, substantiated by hundreds of obser- 

 vations, that the closely adjacent twins of a 

 pair, where the placentas are nearly in contact, 

 are closer duplicates that are individuals taken 

 from the two pairs. 



Chapter V. considers the phenomenon of the 

 free-martin in cattle, or the occurrence of a 

 normal male twin with an imperfect twin, 

 variously considered an hermaphrodite, an im- 

 perfect female, or an imperfect male. The au- 

 thor was fortunately able to avail himself of 

 the work of Lillie and his pupil Miss Chapin, 

 previous to its publication (J. Exp. Zool., 

 July, 1917) and thus presents this work as 

 revealed by the latest investigation. This 

 shows conclusively that the free-martin is a 

 sterile female, with abortive gonads, and with 

 certain of the secondary characters of the male 

 due to the influence of male hormones from 

 the associated male, obtained from the blood 

 circulating in the common placenta. This is 

 a totally different phenomenon from that pre- 

 sented by armadillos, as the twins are here of 

 the fraternal type {dizygotic), and in the 

 latter true duplicates {monozygotic). 



The two final chapters, VI. and VII., show 

 the various contributions to general biological 

 problems afforded by the study of twins, espe- 

 cially in the case of variation and heredity, 

 and here the work of the author and his asso- 

 ciate on armadillos, where the scales of the 

 carapace are used to show the amount of iden- 

 tity, links up extremely well with that of 

 Wilder on human twins, who has employed in 

 a similar way the conformation of the friction- 

 ridges of the palms and soles. Indeed, there is 

 probably more than a general correspondence 

 in method between these two independent 

 series of investigations, since it is altogether 

 likely that the human friction-ridges are 



