R ETROGRESSION. 177 



eater. Or, if we pass in the other direction, we travel 

 quite naturally from the Amphibians to the fishes. In 

 either case, however, in order to complete the circle, one 

 has to get from fishes to Mammals, or vice versa. The 

 gap thus caused is, however, a hopeless one, since the 

 really intermediate groups of the reptiles and the birds 

 have been left on the other side of the circle. 



(2) The second proposition of the circular system is, 

 that 'the primary circular divisions of every group are 

 three actually, or five apparently.' Differences of opinion 

 have arisen among the advocates of the circular classi- 

 fication as to the number of groups in each division, 

 though all were agreed that the system was based upon 

 some fixed number, which governed all the subdivisions 

 from the highest to the lowest. Macleay, the founder of 

 the system, thought there were five main divisions and 

 five smaller ones (making ten in all). Fries, the botanist, 

 adopted four as the number governing his classification 

 of plants; but as he admitted that his central group 

 could always be split into two, this practically made his 

 system likewise quinary. Swainson, as seen in the above- 

 quoted proposition, adopted three as his number; but 

 as one of these groups was supposed to be always 

 divisible into three smaller sections, this also was 

 practically identical with taking five as the ruling 

 number. Other ' circular ists,' again, adopted seven as 

 the governing number. Five was, however, the number 

 most usually adopted, and for this reason this system 

 has often been spoken of as the ' Quinary Classification.' 

 According to this, therefore, everything must go by fives. 

 If animals obstinately refused to range themselves into 



