189IJ 



THE BRITISH NATURALIST. 



229 



The following British land and freshwater shells are all that I know 

 of as possessing a " w. decollatum " : Bythinia tentaadata, Limncea peregva^ 

 L. glabra, Clans, rugosa,. Additions are much to be desiderated. I 

 have had from Lucknow, (India), many specimens of Paludina 

 melanostoma Bars., characteristically decollate, but am not aware of its 

 Occurrence in P. contecta or P. vivipara. 



Brockton Tomlin, Llandaff, 15th August, 1891. 



Bythinia Tentaculata in the Erie Canal. — The Nautilus of last 

 month records that since the introduction of Bythinia tentaadata into 

 the Erie Canal, New York, some years ago, it has driven out the purely 

 American species Goniobasis virginica by consuming their food. 

 According to the experience of Mr. W. M. Beauchamp, who has 

 obseryed the species from its first discovery in alien waters, it prefers 

 canal waters to any other habitat. We hardly think that English 

 observers will confirm this. — Id. 



Vertebrata. 



"WHAT IS A BIRD?" 



BY LINNiEUS GREENING. 



(CoucUided from page 167 .) 



Having seen how birds breed true to their species, subject to those 

 variations which, though common to all groups of organisms, are in 

 birds so slight as to be scarcely perceptible, we can realize how great 

 has been the lapse of time necessary for the production of our present 

 well-marked species and how many centuries have rolled away whilst 

 such slight individual variations have been accumulating till our 

 existing birds have arisen from their far-off reptilian ancestor. 



The fact that birds breed so true to their species is by some made 

 the basis of their their objection to the doctrine of descent. Such 

 objections, however, arise from an imperfect apprehension of the facts. 

 Existing organisms, by their existence, prove their adaptability to 

 their surroundings; unless their surroundings materially alter they 

 will not themselves materially alter. If the struggle for existence be 

 keen, the probabihty of any deviation from the parental type is 

 niimensely diminished, for the simple reason that only favorable 



