PROGRESS IN THE STUDY OF VARIATION. 559 



worn pursuit from which little novelty can be hoped. Such 

 collecting is little better than a pastime, a thing of momen- 

 tary value. On the other hand, by onl)- a slight modifica- 

 tion in his practice the collector could make his collection a 

 body of evidence of high scientific value, evidence which if 

 we are to progress with the problem of species has become 

 a necessity. Not the less will he find in his work the 

 peculiar fascination which collecting exercises over so many 

 minds. 



What the evolutionist asks of the collector is simple. It 

 is, in brief, that he will give to the collection and observa- 

 tion of a few — and for choice, variable — species, the time 

 and effort that he now devotes to the bringing together of 

 rarities. This is a very simple thing, but I believe that at 

 this moment it is a work which scarcely any one has in 

 hand. Scattered over the world there are hundreds of 

 people employing their energies and exercising a high 

 degree of skill in the collection of natural objects and in 

 breeding specim.ens for their cabinets. If only a few of 

 these had set to work in the way required, we should 

 perhaps already know a good deal about the origin of 

 species. 



The fancy is common that in order to investigate evolu- 

 tion, experiments of an elaborate character, dealing with the 

 causes of variation and other recondite subjects, are essen- 

 tial. This is a complete mistake. The first requirement is 

 to know the plain facts as to the occurrence of species and 

 their varieties in the state of nature. To pursue the ques- 

 tion of the interrelations of these forms and the degree to 

 which characters are inherited, artificial breeding must of 

 course be undertaken ; but even such breeding involves in 

 some cases little more than what many of the best collectors 

 — Lepidopterists at all events — are already undertaking for 

 the sake of rearing perfect specimens. 



It may be remarked that variation is conspicuously a 

 subject which can be studied by simple means. Since, 

 moreover, it is essential that each observation should be 

 made repeatedly by different observers in different localities, 

 for a long time no one need fear that he is uselessly 



3S 



