28 



SCIENCE. 



[N. S. Vol. XVI. No. 392. 



now, Papilio be divided into tliree subgenera, 

 Papilio s. str., Cosmodesmus, and Pharma- 

 cophagus, and be still further separated into 

 the African, Indo-Australian, Europo-Siberian, 

 and American forms, it is found that the in- 

 sects of the subgroups still display the tend- 

 ency to have three or nine spots, or bands 

 extending over three or nine interspaces. This 

 is not a matter of correlation, for only 32 of 

 the 453 species of Papilio display hoth three 

 and nine spots upon their fore-wings. It is 

 somewhat difficult to explain this condition ■ 

 upon the hypothesis of natural selection, owing 

 to the fact that Papilios of widely separated 

 regions show the same tendency to produce 

 these two maxima in the same manner. The 

 Hesperidag and Castina show no such tend- 

 ency, hence it is not universal for Lepidop- 

 tera. If it be due to natural selection acting 

 upon Papilios and restricting them to this 

 condition, such selection must be universally 

 operative in the case of Papilio^ but not in the 

 other species. It is easier, therefore, to assume 

 a race tendency in Papilio to produce either 

 three or nine spots upon the fore-wing, or 

 bands extending over three or nine inter- 

 spaces. Other results, quantitatively expressed, 

 were brought out by the author. 



Mr. Trowbridge gave the results of observa- 

 tions on flying birds for the purpose of show- 

 ing that the emarginate primaries of hawks, 

 eagles and certain other birds are interlocked 

 in flight. The speaker referred to his original 

 paper on the subject in which the theory was 

 set forth, which was presented by the late Pro- 

 fessor W. P. Trowbridge before the National 

 Academy of Sciences and the New York Acad- 

 emy of Sciences. The paper created some dis- 

 cussion in SciEXCE at the time, participated in 

 by Dr. Elliot Cones, Professor Newberry, Pro- 

 fessor Trowbridge and others. Mr. Trow- 

 bridge showed by a number of diagrams and 

 photographs that the primary feathers of a 

 number of birds are emarginate near their 

 ends, and that the webs of the feathers are so 

 shaped that when they are overlapped, a 

 curved and rigid aeroplane is formed at the 

 end of the wing, which, he considered, is of 

 considerable advantage in swift sailing flight. 

 The emarginations of the primaries of hawks 



and eagles are particularly pronounced, and 

 permit firm interlocking. A table of observa- 

 tions was given, showing that the interlocking 

 of the primaries does take place, the data 

 having been obtained at New Haven during 

 the autumn flights of hawks along the Con- 

 necticut coast. It appears that in the ease of 

 one species of hawk examined, ten wings out 

 of forty had all five primaries interlocked, and 

 that the number of wings having sixty per 

 cent, of the primaries interlocked was twenty- 

 nine, or 72 per cent, of the total number, forty. 

 It was concluded that emarginate primaries 

 of hawks and other birds are interlocked in 

 flight on the following grounds: (1) it has 

 l)een found that the webs of such feathers of 

 hawks that had just been killed usually show 

 deep notches where they have rested against 

 one another, which notches could only result 

 from habitual interlocking of the primaries; 

 and (2), in every case of over 25 hawks 

 killed while flying and examined immediately 

 after they fell, some prirnaries were inter- 

 locked (several slightly wounded birds not in- 

 cluded). In the case of 19 perfect specimens 

 cif one species, 67.9 per cent, of all emarginate 

 primaries (190) were found to be interlocked. 

 'WTiile it is not possible at present to show 

 when the emarginate primaries are interlocked 

 in flight, the indications are, however, that this 

 occurs when the wing is partly flexed, as in 

 the case of hawks sailing rapidly througii 

 the woods and flying in a strong wind. The 

 important functions of interlocking appear to 

 be (1) to make more rigid the outer portion of 

 the wing, that part of the aeroplane formed by 

 the primaries, and (2) to produce a curve of 

 the wing which enables the bird to have a bet- 

 ter control of its swift flight through the air 

 than the unlocked condition would permit. 

 The end of the bird's wing when the primaries 

 are interlocked becomes shaped somewhat like 

 the blade of a propeller screw. The interlock- 

 ing also would keep the primaries extended 

 without muscular exertion on the part of the 

 bird. 



Considerable discussion was arotxsed by Mr. 

 Trowbridge's paper. Dr. Jonathan Dwight, 

 Jr., presented a series of arguments against 

 the theory of the speaker, to the effect, in 



