432 



SCIENCE. 



[N. S. Vol. XXII. No. 5G2. 



In view of the foregoing facts, I am of tlie 

 opinion that in H. lecubronchialis and. in H. 

 lawrencei, . . . we have examples of two sepa- 

 rate and distinct ' mutations ' from a common 

 parent stock or species. Tliat is, I believe that 

 H. pinus, early in the last centviry became un- 

 stable as a species and began to throw off what 

 must be considered as ' mutants,' taking de 

 Vries's definition of the word. In other words, 

 H. pinus is alone responsible and is the direct 

 ancestor of both H. leucobronchialis and H. law- 

 rencei; that these ' mutants ' have up to the pres- 

 ent time generally bred back into the parent 

 stock, and that in so doing the instability of H. 

 pinus has increased geometrically with the con- 

 stant result of the increasing number of both 

 kinds of ' mutants.' 



While the ' mutation ' theory may be a 

 good hypothesis to consider in respect to these 

 peculiarly unstable groups, of birds, it must 

 be noted that the method of their origin and 

 the results, as now known, are very unlike the 

 methods and results of mutation in plants, as 

 made known by de Vries. The facts and con- 

 ditions are not to any great extent parallel. 

 Instead of the resulting ' mutants ' remaining 

 constant and breeding true, as in the case of 

 primroses, they are in this case unstable and are 

 believed' to interbreed freely with each other 

 and the parent stock. Besides, in building up 

 his theory of ' mutants ' in the case of these 

 warblers, we think Mr. Scott has belittled the 

 evidence of hybridity and laid too much stress 

 upon the (assumed) completeness of knowledge 

 ' in the early part of the last century ' of the 

 ornithology of the area now inhabited by these 

 birds. While it is true that most of these 

 puzzling birds have been taken within the last 

 twenty or twenty-five years, it does not follow 

 that, as Mr. Scott says: 



It is not likely that a form or kind of bird so 

 common as H. leucoironchialis is at the present 



^ By those who are most familiar with the facts. 

 Interbreeding is known to occur between the two 

 stock species, and also between their offspring and 

 both of the stock species, and it has been re- 

 peatedly assumed by the best authorities that the 

 hybrids are fertile infer se. This feature of the 

 case is of course impossible of demonstration, 

 owing to the nature of the conditions — the im- 

 possibility of continued observation of the same 

 individuals for a series of years. 



time, and ranging over as large an area as from 

 Pennsylvania to Massachusetts and from Virginia 

 to Michigan, should remain unknown to the earlier 

 ornithologists, such keen field naturalists as 

 Audubon and Wilson, Baird, Lawrence, Coues and 

 Prentiss. Nuttall made careful and prolonged 

 study of birds in the region where Mr. Brewster 

 collected the type. Yet none of these close ob- 

 servers and good collectors either recorded or 

 secured an individual of this kind. Clearly then, 

 the presumption is that this bird could not have 

 been so common early in the last century as it is 

 now, if indeed it existed at all at that time. 



Eirst as to the range of these two forms, 

 with reference to that given in the above 

 quotation. H. lawrencei has been found only 

 in the northern part of New Jersey, the lower 

 Hudson Valley, and the lower Connecticut 

 Valley. li. leucobronchialis has but five 

 records (all of migrants) south of northern 

 New Jersey, two of which are for southeastern 

 Pennsylvania, two for the immediate vicinity 

 of Washington, where collectors abound, and 

 the other (not mentioned by Scott) for 

 Louisiana. The bird has been reported as 

 observed in northern Ohio, but the only record 

 of a captured specimen for the region west of 

 New Jersey and eastern New York is a single 

 bird taken in southern Michigan. There are 

 also only two records for the region north of 

 Connecticut, which include the original type 

 specimen (Newtonville, Mass., 1870) and one 

 other (Hudson, Mass., 1858). Thus the known 

 distribution of these forms, at least for the 

 breeding season, is narrowed down to prac- 

 tically northern New Jersey, the southeast 

 corner of New York (extreme lower Hudson 

 Valley) and Connecticut. This is quite dif- 

 ferent from the distribution conditions that 

 might be implied from the sweeping statement 

 above quoted from Mr. Scott. 



Now as to the work of the earlier natural- 

 ists. Both Wilson and Audubon explored the 

 region around Philadelphia, where, notwith- 

 standing all the careful field work of many 

 expert collectors during recent years, there 

 are only two records for leucoironchialis and 

 none for lawrencei. These naturalists also 

 each made journeys to New England, but their 

 visits were brief and for the most part with 

 other interests than field work, and it is well- 



