( Ixxxv ) 



species to take diagnostic rank, then the systematist may 

 pursue his labours with renewed zest, inspired by the idea 

 that there is something more underlying his studies than the 

 association of a name with a " type." 



How far the principle of physiological correlation will 

 enable us to explain trivial and non-significant characters 

 I will not pretend to say. I merely formulate the sugges- 

 tion in order to prompt observation and experiment, and I 

 ask systematists to allow that indirect utility of this kind 

 may have to be reckoned with. In the interpretation of 

 diagnostic characters we have to consider that some are 

 obviously useful now, that others may be ancestral, that 

 others are obvious correlates, that others, again, must have 

 some direct use of which we are ignorant, and that a residue 

 of inexplicable characters may exist of which some are the 

 correlates of latent specific characters. Considering the 

 small amount of progress that has been made in the physio- 

 logy of the lower classes of animals, it is not surprising that 

 I should have been unable to draw largely upon entomo- 

 logical examples. It is certainly to me a matter of wonder 

 that the species question should not have been handled to 

 a greater extent by professed physiologists. Most cordially 

 can I endorse the recent utterance of Prof. Ludwig von 

 Grafl": — 



" Looking back, we see how in all the chief branches of 

 zoological science the theory of descent newly formulated by 

 Darwin has become the motive of a thoroughness in research 

 not found in any earlier period. It is characterised by the 

 preponderance of the morphological interest, which has led to 

 such a one-sided neglect of physiology, that to-day, when the 

 development of morphology forces the formulation of questions 

 whose answers experiment alone can supply, neither the 

 methods of work nor the worker himself are at hand to solve 

 them."* 



Although in the absence of a sufficient body of definite 

 experimental evidence physiological correlation can only be 

 regarded as a hypothetical factor in the production of specific 

 characters, it has appeared to me sufficiently plausible as a 



* "Natural Science," Vol. IX., p. 368. 



