1888.] 



possibly Dual Origin of the Mammalia. 



377 



other mammals,* the one being formed from modified sweat glands, 

 and the other from sebaceous follicles. If this distinction is found to 

 hold good throughout the class, it seems to me difficult to think that 

 the Mammalia had not this dual origin — an hypothesis which har- 

 monises so well with the differences, skeletal, genito-urinary, and 

 developmental, which divide these two groups of mammals. 



On this view, the teeth of the toothed Ornithorhynchus ancestor 

 must have arisen for the first time in a form more reptilian than is 

 the form of our living Monotremes, yet sufficiently divergent from 

 the Sauropsidan main stem to explain the non-existence of teeth of 

 the kind in any known Sauropsidan, living or fossil. 



To this hypothesis it will probably be at once objected, that 

 Mr. Caldwell'sf studies of the mammalian ova show a noteworthy 

 resemblance between those of the Marsupials and Monotremes. But 

 if the Marsupials are an offshoot from- the placental mammals, then 

 such resemblances as exist between them and Monotremes in this 

 respect must be induced resemblances. Moreover, certain very note- 

 worthy resemblances exist between the ova of those exceptional 

 Amphibians, the Ophiomorpha, and Sauropsidan ova. J It may be 

 objected in the second place that the dual hypothesis implies the 

 independent origin of too many similar structures. But the inde- 

 pendent origin of similar structures is a doctrine for which I have 

 combated ever since the year 1869. I say " similar," not "identical." 

 No two leaves in a forest are absolutely alike ; how then could absolute 

 resemblance be thought possible between two structures of different 

 origin ? Yet the closeness of resemblances between parts which must 

 have arisen diversely is often remarkable. The Marsupials are now 

 regarded as having diverged from the mammalian stem by some single 

 remote ancestor. Yet amongst its descendants have arisen animals 

 some of the teeth of which strikingly resemble some of the teeth of 

 beasts of the placental series. Some teeth of Perameles and TJro- 

 trichus, of Macropus and Macroscelides, of Thylacinus and of Canis, 

 may be cited as examples ; and though the histological difference of 

 the extension of dentinal tubes into the enamel generally obtains in 

 the Marsupials, yet it is more marked in the Kangaroos, which are 

 the most differentiated forms, while such tubes almost or quite vanish 

 in the Dasyuridne, which more nearly resemble ordinary mammals. 

 But the most striking similarity of tooth structure is that between 

 Orycteropus and Myliobates — a similarity which extends over the micro- 

 scopic characters. Again, it would be difficult to find a more curious 

 practical resemblance than that between the hinge teeth of Lophius, the 



* See his ' Zur Kenntniss der Mammarorgane der Monotremen,' 1886. 

 f < Phil. Trans.,' B, vol. 178 (1887) p. 463. 



X See the account of the ova of Ichthyophis fflutinosvs in C. and P. Sarasin's 

 1 Ergebnisse Naturwiss Forschungen auf Ceylon,' vol. 2, 1887, p. IX. 



