230 



SCIENCE. 



[N. S. Vol. XVI. No. 397. 



To many persons it would seem to be almost 

 an axiom that a character which can not be 

 stated in language or in figures of any sort is 

 not sufficiently conspicuous to bear the weight 

 of a name. But Mr. Oberholser, holding a 

 point of view occupied by many ornithologists 

 and mammalogists (and perhaps other zool- 

 ogists) which is adding to current zoological 

 literature hundreds, if not thousands, of 

 trinomials every year, thinks differently; he 

 says (page 803) : " Various more or less per- 

 fect intermediates are very perplexing, and no 

 means of determination can possibly be of 

 value except the actual comparison of speci- 

 mens, coupled with an accurate knowledge of 

 the relative value of the proper differential 

 characters. Satisfactorily to present such in- 

 formation in printed diagnoses is manifestly 

 out of the question, for characters that will 

 serve to identify even typical examples of some 

 of the more closely allied forms are frequently 

 almost impossible to express intelligibly on 

 paper." Notice especially the contention that 

 even the actual comparison of specimens is not 

 itself sufficient for the identification of a "bird 

 unless such comparison is made by an expert. 

 If this is so, systematic ornithology is in a bad 

 way, for if the expert can not express the dis- 

 tinguishing characters ' intelligibly on paper,' 

 what are we going to do when he dies? Of 

 what possible use is it to attempt to maintain 

 distinctions so fine that even a well-trained 

 ornithologist can not tell upon what form his 

 observations are made? Such distinctions 

 might be of value if they had any geographical 

 meaning, but even this is denied them in the 

 ease of Otocoris, for in Kansas the student 

 must distinguish between four possibilities, 

 and in northwestern Mexico near the interna- 

 tional boundary line, he may come upon any 

 one of seven of Mr. Oberholser's ' subspecies.' 

 Nor will the season of the year help him much, 

 for in Kansas at least three forms occur in 

 winter, and in northwestern Mexico near the 

 ' Line,' unless the larks keep very strictly to 

 the limits laid down for them, no less than half 

 a dozen forms may breed. 



Another rule which to the layman would 

 seem to be axiomatic is that characters which 

 can not be recognized regardless of the locality 



where the specimens are collected are worth- 

 less. ButJMr. Oberholser says (page 803) that 

 ' the identification of specimens without regard 

 to geography is, to say the least, liable to be 

 difficult.' No one can read the paper carefully 

 and not realize the magnitude of the difficultj^. 

 An illustration may be taken from the sub- 

 si^ecies chrysolmma, which is given as resident 

 in Mexico. Speaking of some specimens from 

 Puebla and Vera Cruz, this statement is made 

 (page 844) : " If comi>arison be instituted 

 between these specimens and typical actia 

 from California, however, it will be at once 

 seen that they are exceedingly similar, and, to 

 say the least, difficult to distinguish, forming 

 another of those perplexing cases of forms 

 reduplicated by apparent intergradation of two 

 or more others." In other words, these birds 

 are 0. a. chrysolwma in Mexico, but if taken 

 in southern California, they would be 0. a. 

 artia! 



Another point which will be a great surprise 

 to many unsophisticated persons is the recog- 

 nition given to very slight differences in size. 

 Many examples might be given, but the follow- 

 ing will suffice. 



0. a. enthymia is said to be ' decidedly 

 smaller ' than 0. a. arcticola, but the latter has 

 the wing (total lengths are not given) averag- 

 ing only 6.7 mm. longer than the former, the 

 tail 0.3 mm. and the exposed culmen is the 

 same in both forms. Even the apparent dif- 

 ference in the length of wing is not really so 

 great, for only fifteen specimens of each form 

 were measured, and the difference between the 

 maximum and minimum wing measurement 

 of the two is just 1 mm. Again the subspecies 

 peregrina is based on a single specimen (from 

 Colombia), which is distinguished from insu- 

 laris (from the Santa Barbara Islands) only 

 by size. It is said to be ' very much smaller.' 

 As a matter of fact, the wing is only 5 nun. 

 shorter than the average insularis, the tail 1.4 

 mm. shorter and the culmen .08 mm., and, 

 moreover, the tail is 3 mm. longer than the 

 minimum insularis, and the bill, tarsus and 

 middle toe are each half a mm. longer than the 

 minimum for that species. Clearly we have 

 here an exaggerated idea of the length of a 

 millimeter; the 'inch on the end of a man's 



