312 COOK 



can start a new race or variety, but with sexually differentiated 

 animals this is much more difficult, since interbreeding is neces- 

 sary for reproduction. An actual instance will illustrate the 

 point. In all the millipedes of the world-wide order Merocheta 

 the olfactory cones of the antennae are four in number, arranged 

 in a square, with the single exception of a series of closely 

 related East African genera of the family Gomphodesmidae, 1 

 which are unique in the possession of ten olfactory cones 

 arranged in a circle. That the four cones in a square is the 

 ancestral condition, is certain, because it is shared also by all 

 the other orders of the very ancient class. Diplopoda, many 

 members of which are known from the carboniferous period. 

 That the number is invariable in the order Merocheta can not 

 be claimed, since, obviously, it must have varied at least once, 

 when the circle of ten cones came into existence. No variation 

 has been recorded, however, either in the four-coned or the ten- 

 coned genera, on the many thousands of specimens which have 

 been examined. 



Nor are there any indications that the ten-coned condition is 

 an advantage which has gained any favors from natural or other 

 forms of selection. The ten-coned genera as a group show no 

 other conspicuous peculiarity and have contributed, apparently, 

 only m average share to the evolutionary diversification and 

 geographical distribution of the family. Moreover, the habits 

 and environmental relations of the whole class Diplopoda are 

 such as to reduce the influence of natural selection to a 

 minimum. 2 



Under such circumstances the sidewise origination and pres- 

 ervation of a ten-coned new species as a mutation seems highly 

 improbable, but there is, on the other hand, no reason why 

 a genetic variation to ten cones should not spread through a 

 species and be carried forward into the other species and genera 

 into which the ten-coned group might afterward subdivide. If 

 there had ever been millipedes with the intervening number of 



1 Cook, O. F., 1899. African Diplopoda of the Family Gomphodesmidae. 

 Proc. U. S. National Museum. 21 : 677-739. 



2 Cook, O. F., 1902. Evolutionary Inferences from the Diplopoda. Proc. 

 Entomological Society of Washington. 5 : 14. 



