BY R. M. JOHXSTOX, F.L.S. 37 



of differing environments are unknown or obscure, so long 

 must we be dissatisfied with a classification which so largely 

 depends on the theory of hybridism to account for the vast 

 number of intermediate forms which link together the several 

 closely-allied types, now artificially erected into species for 

 the mere convenience of local classification. 



These remarks ai*e not intended to reflect upon the necessary 

 classification adopted locally for museum collections. They 

 are only intended as a jjrotest against the classification so 

 artificially based when it is assumed to be in truth naturally 

 fixed, and capable of maintaining the various characters 

 unmodified by transference to the widely changed conditions of 

 a new environment ; as for example, the transfer of selected 

 types of European sjiecies to the waters of Tasmania. 



When the few trival distinctions which alone serve to 

 support the adopted nonemclature of Europe fail to appear 

 in what in all probability are deemed to be the true acclima- 

 tised descendants of such species, we have no right to 

 assume upon such uncertain ground that the characters of 

 their descendants are so fixed as to remain unaffected by tbe 

 new conditions lander which they live. It is quite jiossible 

 that it may be so ; but that is an open question. That they are 

 not so fixed is at least equally possible ; and this conception, 

 moreover, is more pi'obable when all tlie facts of the case are 

 taken into consideration. When individuals show one or 

 two peculiar characters in one environmeut which are not 

 reproduced by what appears on good evidence to be their des- 

 cendants in another widely differing environment, it is more 

 reasonable to assume that the characters have been modified 

 by the transfer, than that the extreme forms so largely and 

 successfully introduced into our waters should altogether 

 cease to exist, or vanish from our shores. It must be borne 

 in mind that among fishes showing every gradation of change 

 within the limits of variability, the predominant tyjjes in one 

 locality may be due to the influence of local environment, 

 rather than to hereditary influences. To assume, as is too 

 frequently the case, that such prevailing types indicate 

 greater purity of breed, is to beg the whole question at issue. 

 It is well known that the prevailing forms of sea-trout in 

 English, Welsh, and Scotch streams, dift'er so considerably 

 with the locality that classifiers regard them as distinct 

 species. The forms known as S. trutta, S. (jellivensis, S. 

 Camhricus, S. hrachyijoma, are examples of this class. 



But although the minor characteristics which served im- 

 perfectly to distinguish these types are admitted, there is 

 no proof that the prevalent type characters are not purely the 

 effect of local environment which miglit be speedily obliterated 

 or transformed by transfer to a different environment. The 



