44 



The Canadian Field-Naturalist. 



[Vol. XX^'. 



coloured head of the female hrooksi does 

 really enable us to separate the two races 

 with some confidence. The bill averages a 

 little slenderer in hrooksi than in eastern 

 specimens, but this is not a reliable guide, 

 for the difference is so slight that it is 

 hardly recognizable even in the figures that 

 accompany- Grinnell's description, and at 

 times is quite swamped by individual 

 variation. 



I have been fortunate in having sub- 

 mitted to me for examination the two birds 

 that were taken b;^- RoAvan at Indian Bay. 

 One is an adult female in rather worn and 

 faded breeding plumage, the other her 

 fully fledged young still suffused on back 

 and breast ^ith the ochraceous buff that 

 is lost in the post-juvenile moult before 

 the first winter. There can be no ques- 

 tion that the latter is a bird of the year, 

 and the fact that it was still being fed by 

 its parent indicates that it was raised with- 

 in a short distance of where it was taken. 

 The parent that was captured is fortun- 

 ately a female, and consequently of the sex 

 needed for subspecific determination, but 

 the disturbing fact is that it bears the 

 colour marks of the western bird, hrooksi, 

 rather than those of the expected eastern 

 vespertina. Its head, in contrast with its 

 back, is decidedly dark, and the specimen 

 generally, in spite of its slightly worn and 

 faded condition, is identical with birds 

 from the Okanagan Valley, B.C. It is true 

 that the bill is large and stubby rather 

 than attenuated, and that a large stubby 

 bill is a postulated character of the east- 

 ern bird; but, as mentioned before, this, 

 ir the opinion of the writer at least, is 

 too variable a character to be of much 

 service in determining individual speci- 

 mens. Of course, it does not seem rea- 

 sonable that the British Columbian Even- 

 ing Grosbeak, hrooksi, breeds eastward to 

 Lake Superior nor would I care to suggest 

 such a conclusion except to negative its 

 probability. We may prefer to regard 

 this specimen simply as a variant of the 

 ea.stern vespertina, or perhaps suspend 

 judgment until further evidence is secured. 

 This unfortunate occurrence of abnor- 

 mality in a unique specimen brings up the 

 important subject of the determination of 

 slightly fharacterized races when thev are 

 found far from their natural habitats. 



Birds with their great mobility certainly 

 can and do wander to the most astonishing 

 localities, and it is not impossible for oc- 

 casional specimens to appear far from the 

 land of their origin and direct blood re- 

 latives. In such cases are we to identify 

 entirely from the characters that the birds 

 exhibit? If we do, we are bound to make 

 a multitude of errors through mistaking 

 individual for racial variations. On the 

 other hand if we allow considerations of 

 geography to influence our identifications 

 we are just as certain often to twist the 

 evidence to suit our geographical precon- 

 ceptions. 



There are good reasons for either course. 

 If a race means anything it must be based 

 on germinal characters and denote blood 

 relationship between the individuals com- 

 posing it. An albino Negro would not be 

 a Caucasian, however close the superficial 

 resemblance might be. The very fact that 

 an isolated community of a species has de- 

 veloped certain common characteristics in- 

 dicates the possibility, perhaps even sug- 

 gests an innate tendency, of the species as 

 a whole to vary in that special direction ; it 

 would surprise us less to find such a var- 

 iation sporadically in individuals of other 

 communities than one the possibility of 

 which has not been demonstrated. In 

 other words, we can expect to find, even 

 in pure lines of descent, disturbing variants 

 (sports, if you Avill) resembling estab- 

 lished races more often than departures in 

 novel directions. 



These considerations are against identi- 

 fying by character alone without consider- 

 ing geography as an indication of probable 

 descent. However, to lay too great stress 

 on geography is equally dangerous and mis- 

 leading. For if we plot di.stribution on the 

 determination of specimens we certainly 

 must not make postulated distribution, the 

 basis for such determination or we shall 

 be reasoning in a vicious circle. All we 

 can do in doubtful cases is to acknowledge 

 our ignorance, and content ourselves with 

 naming the species, leaving the determina- 

 tion of the sul)species open for further in- 

 vestigation or fuller data. 



In this case of the Evening Grosbeak, 

 then, although we may be firmly convinced 

 that the Indian Bay breeding bird i.s of 



