WHAT IS A SPECIES? 261 



may not have arisen from a distinct creative act, and thus the 

 term species here too becomes a more or less arbitrary division 

 and dependent on each man's phantasy. 



Since then whatever view we take of the scheme of creation, 

 species must, at any rate until our present knowledge is infinitely 

 multiplied, still remain an arbitrary term, let us endeavour so 

 to define it as to leave as few doubts as possible as to the exact 

 limits we arbitrarily assign to it, so that it shall be moderately 

 certain that referring to a common standard, what one naturalist 

 admits as a species shall be admitted by others, and that what 

 one rejects others also shall as a rule reject. 



A simple illustration will explain perhaps my view of our 

 position in regard to this question. 



If we look at a map of the Indian Archipelago, with its com- 

 plicated groups and clusters of thousands of islands, we may 

 suppose either that each of these indicates a separate centre of 

 upheaval, or that these are all that is left to us in a wide area 

 of subsidence or submergence of what was once a compact 

 continent ; or we may suppose them to be the results partly of 

 elevations, partly of depressions in the crust of the earth, and 

 partly of alterations in the oceanic level independent of these ; 

 but adopt what hypothesis we will as to their formation, this 

 ought surely not to affect our decision as to the method of 

 entering on that map the names that are to facilitate our refer- 

 ence to the various places and the record of our observations in 

 regard to each. 



We may give each separate island big and little a distinctive 

 name, we may group little clusters under one name, and give 

 long promontories of the big islands separate names ; it is all a 

 matter of convenience, and cannot and ought not to be governed 

 in any way by our conceptions as to the manner in which the 

 existing state of things has eventuated. 



For my part, for the sake of obtaining a simple and intelli- 

 gible rule, that all who run may read, I would give every 

 island, however small, divided off by ever so narrow a channel, 

 a separate name ; but I would give no separate name to a pro- 

 montory,* however long. 



I would therefore adopt the popular definition of a species 

 with which I commenced, and would further define an essential 

 difference to be one however small that is constant and that is 

 not bridged over by intermediate links. 



This latter clause appears to be essential. So lono- as the 

 promontory remains attached, I should deprecate bestowing on 

 it a distinct name as if it was a separate island. Where the 



* Some acute individual will enquire what I would do with a place that is an island 

 at high and a promontory at low water. I can only reply that as it is not always an 

 island, I shall treat it as a promontory. 



