134 THE SURVIVAL OF THE UNLIKE. [iV. 



forms which we, merely for purposes of convenience in 

 classifying our knowledge of them, combine into more 

 or less marked aggregations, to which we have given the 

 name species. Very often we find in nature an aggre- 

 gation of successive individuals which is so well marked 

 and set off from its associated groups that we consider 

 nature to have made an out and out distinct species; but 

 a closer acquaintance with such species shows that, in 

 many cases, the intermediate or outlj'ing forms have 

 been lost, and that the type which we now know is the 

 remainder in a continuous problem of subtraction. In 

 other cases, a form appears to have arisen without inter- 

 mediate forms, as a distinct offshoot from an older type. 

 This is well illustrated in many remarkably distinct 

 garden forms, which originated all at once with charac- 

 ters new to the species or even to the genus. I have 

 mentioned such a case in the Upright tomato. Even the 

 sudden appearance of these strange forms is proof that 

 species may originate at any time, and that it can be no 

 part of our fundamental conception of a species that it 

 shall have originated in some remote epoch. Species - 

 making forever enforces the idea of the distinctness and 

 immutability of organic forms, but study of organisms 

 themselves forever enforces an opposite couf^eption.* 

 The intermediate and variable forms are perplexities 



* If this position is well taken, it follows that the naturalist should not describe 

 new species with the idea of adding anotlier item or organism to the inventory of 

 nature, but for the purpose of classifying and clarifying our knowledge of the 

 kind and extent of variation which the given group presents. A new species, 

 therefore, is made simply for convenience's sake. In very variable groups, it is 

 perfectly justifiable to mjike species when it is known tliat occasional forms are 

 intermediates, if thereby we are enabled to understand the relationships of the 

 various forms more clearly. This is particularly true in narrow groups which 

 have many forms of varying taxonomic importance. An illustration may be taken 

 from the genus Carex. The echinatu group contains four more or less coordinate 

 main types, the echinata proper of the Old World, and three types in the United 



