244 THE DISTRIBUTION AND CHARACTERS OF REPTILES. 



isolation may have played an important part : the former occurs 

 in the Transvaal, Free State, and Bechuanaland, the latter in the 

 carroid areas of Cape Colony : it should be added that the differ- 

 ences between them are very slight, being based mainly on the 

 degree of keeling of the dorsal scales. 



The fact that many genera can be structurally divided into 

 groups of closely related species, the members of each group 

 being geographically separated whilst members of different groups 

 may live together in the same locality, may imply 

 that species formation of a larger kind occurs without 

 the help of geographical isolation : for instance the species of the 

 geometric group of Tcstndo have a geographical significance, but 

 besides these the genus includes some very distinct species, such 

 as the giant tortoise {T. pardalis) which commonly occurs in the 

 same localities as any of the geometric tortoises, and there is no 

 reason for supposing that the giant and geometric tortoises ever 

 were geographically separated. 



Mutations. — By many biologists the mutation theory has 

 been thought to afford an explanation of the discontinuity between 

 species, and saltatory evolution has been welcomed as affording a 

 much better handle for the tool of natural selection than was 

 available in terms of the older Darwinian hypothesis. The term 

 mutation, as used by De Vries, implies a genetic variation which 

 is constantly transmitted to the offspring, and which introduces 

 a new character to the organism : as such it is distinguished from 

 f!uctuational variations, which are not transmissible as definite 

 single characters. Whether these two classes of variation are 

 rigidly distinct may well be doubted, and it is obvious that if we 

 distinguish between them only on the basis of their transmissi- 

 bility, the term mutation might be used to include variations so 

 small as to be scarcely perce]itible, in which case the theory loses 

 much of its value as an explanation of discontinuity or of the 

 origin of new characters. Broadly speaking, therefore, a muta- 

 tion is understood to mean a variation which is relatively large 

 as well as definite and transmissible, and in this sense I employ 

 the term here. There can be no doubt about the reality of muta- 

 tions in certain cases, but that such mutations have contributed 

 to any considerable extent in the formation of species is not sO' 

 certain. The phenomena of Mendelian inheritance include the 

 production of what may rightly be called mutations in the reces- 

 sives, and instances of Mendelian inheritance are common enough, 

 both in animals and plants An interesting case is that of our 

 common rat Mus rattus, of which two colour varieties occur in 

 Grahamstown and elsewhere, a pale-bellied form and a dark- 

 bellied form : these forms have been shown by Mr. Bonhote to 

 follow the law of simple Mendelian inheritance. On the other 

 hand, it has been suggested that the various mimetic forms of 

 the same species of butterfly are of Mendelian origin, but Prof. 

 Poulton has recently shown that in some cases at least such is not 



