THE EMBRYOLOGY OF THE OPOSSUM 209 



Bensley ('Ola) has also pointed out the following facts about the 

 mammary apparatus and the pouch. The observation of Gegenbaur 

 that the mammary glands of monotremes are specialized sudoriparous 

 glands whereas those of marsupials and placentals are of sebaceous 

 origin is sometimes cited as evidence that the marsupials are de- 

 rived from the Eutheria, not from the Prototheria. Bensley quotes 

 Gegenbaur (1880, p. 35) to the effect that he found some sebaceous 

 milk glands even in the monotremes and he recognized the probability 

 that in the original prototheria both sorts of glands were present. 

 In the modern monotreme one type has been emphasized, so to speak, 

 and in higher mammals the other. Klaatsch (1892) found both sorts 

 present in the artiodactyl ungulates, which confirmed this in- 

 terpretation. 



Klaatsch also recognized the homology of the tubular teat cavities 

 of the artiodactyl ungulates and the circular folds which surround the 

 nipples in marsupial embryos. He was of the opinion that the pouch 

 of the marsupials is formed by the fusion of these teat cavities, and 

 correspondingly he argued that the ungulates could not have passed 

 through a marsupial stage. Bensley mentions that this is beside 

 the point, as Huxley had regarded the pouch as a character of modern 

 marsupials, and had suggested that the Eutheria were derived, not 

 from the marsupials, but from the primitive Metatheria. Further- 

 more I am dubious about these teat cavities having anything to do 

 with the formation of the pouch. If the theory of pouch formation 

 that I suggested in a previous section of this paper be correct, then 

 the primitive Metatheria must have had paired, longitudinal, labio- 

 scrotal folds, and from these are derived both the pouch of the 

 marsupials and the labia majora of the placental mammals. In this 

 feature, as in most others, the modern Didelphyidae and Dasyuridae 

 would be nearer the original ancestral stock than any other living 

 marsupials. 



Lest in this emphasis on a few debatable points we lose sight of 

 the mjTiad anatomical features in which the marsupials are nearer 

 the Sauropsida than are the placental mammals let me call to mind 

 again : the presence of both ducts of Cuvier ; the absence of a septum 

 secundum in the heart ; the absence of a corpus callosum between the 

 cerebral hemispheres; the separate squamosal, petrosal, and tym- 

 panic bones; the presence of a complete coracoid cartilage in the 

 embryo. And then we may add a few which have not been mentioned 

 before in this paper. 



The cells of Paneth are not confined to the glands of Lieberkiihn, 

 but are distributed all over the surface epithelium of the intestine, 

 and are primitive in structure. This is like the condition in the 

 lizard, and unlike that of placental mammals (Klein, '06). The 



