NATTJKE OF PARALLELISM. 47 



development ; and as sucli parallelism in development, or 

 shortly "parallelism," lias recently attracted a good deal 

 of attention among biologists, we propose in this chapter 

 to present to our readers some of the more striking 

 instances of this feature. In the instances above alluded 

 to, the parallelism is either to a great extent shown in 

 external characters, or in structures which are easily 

 modified ; but, as we shall indicate in the secjuel, in other 

 cases it atfects deeply-seated structures ; and its inducing 

 cause is then very diflScult to surmise on any of the 

 ordinarily accepted doctrines of evolution. It will, more- 

 over, be obvious that the accejrtation of parallelism in 

 development — and accepted to a certain extent it must 

 undoubtedly be — throws a new difficulty in the inter- 

 pretation of the affinities of animals, since before saying 

 that an identity in some structural feature between the 

 members of any two groups indicates their relationship, 

 we have first of all to determine whether such similarity 

 of structure is due to parallelism, or is inherited from a 

 common ancestral type. As is so generally the case when 

 any new theoi-y is started, a host of enthusiastic writers 

 have welcomed parallelism with acclamation, and have 

 attempted to apply it in a number of instances where 

 there is not at present sufficient evidence of its existence. 

 We have been told, for instance, that the American 

 monkeys have no relationship) with their reputed cousins 

 of the Old World ; that whalebione whales are not allied 

 to the sperm-whale and dolphins ; that cats have no 

 kinship with civets or other modern carnivores ; and that 

 the egg-laying mammals have been evolved from a reptilian 

 or amphibian stock, c[uite independently of all other 

 members of the mammalian class — their resemblances 

 being solely due to parallelism. At present, we confess, 

 we are totally unable to accept any of these conclusions, 

 and we think it somewhat improbable that if the members 

 of either of the above-mentioned pairs of groups had an 

 independent origin, they would have presented such a 

 similarity, both externally and internally, as we find to Ije 

 the case. Still, however, it must be l>orne in mind that 

 there is a considerable amount of evidence that the modern 



