on Sperniophila as a Potential Weaver.



89



structural modifications of beak, foot, and wing, in his search

after the relationships of genera, but rather for those who con¬

sider that the habits of a bird may offer a surer indication of its

affinities than the outline of beak, length and proportions of

tarsus or claws, of wing or tail, (all of which doubtless are

modified by the habits). Without a perfect knowledge of the

internal anatomy, I do not see how anyone can state authorita¬

tively that any external character is inherited and not an adapta¬

tion to the modus vivendi.


In the twelfth volume of the “Catalogue of Birds” Dr.

Sharpe places Spermophila near the end of the Grosbeaks ( Cocco -

thraustince) and near to Melopyrrha and Phouipara , but in his

introductory notes to the Fringillidce he expresses the opinion

that “ Phonipara and Volatinia are probably Buntings.”


Ridgway (Birds of North and Middle America) appears to

me to regard the three Sub-families of Sharpe’s classification as

purely artificial, and their exponent admitted that they were so ;

and owning his own inability to define these sub-families, he

enumerates certain genera in which the young are streaked and

others in which they are not, rejects four genera as not fitting

into any of the groups susceptible to more or less exact defini¬

tion, and concludes with an admission that his key to the genera

is to a large extent artificial.


The classification of any group of animals by external

characters alone must always be to a great extent artificial. A

natural classification must be based not only on the internal as

well as external structure, but also upon the habits of the living

creatures. I note that Ridgway does not ignore this fact, although

he puts habits last: I should be inclined to regard them (within

certain limits) as of greater importance than external structure,

especially where they deal with nidification.


The form in which a bird habitually builds its nest must

have been the result of a habit slowly acquired through many

long forgotten generations of ancestors ; therefore the structure,

especially of complex nests, seems to me a good guide, within

reasonable limits, to relationship. I should not, of course, re¬

gard Phylloscopus as related to Munia because there is a similarity

in the general character of their nests ; but where, among the



