572 PTEEYLOGBAPHY OF GOATSUCKERS AND OWLS— CLARK, vol.xvii. 



seems to have brauelied off from tlie Caprimulgiue form very early and 

 probably by becoming diurnal to a greater degree. Thus having lit- 

 tle need of sensitive rictal bristles, it lost them, while the wing and 

 tail were also modified. Among the owls the variation from the sniJ- 

 posed ancestral form has been more spasmodic and the direct progress 

 much less, but the pterylography of even the American forms is too lit- 

 tle known to draw any satisfactory conclusions. It is certainly a very 

 curious fact that Strix shows some variations which are comxDletely 

 parallel to those of Chordeilcs;^ thus, the outer pair of tail feathers is 

 the longest, the tenth primary almost equals the ninth, the peculiar 

 forking of the lower cervical tract gives a hint of the origin of the 

 inner branch of Chordeiles^ and, finally, the parvial fusion of the sternal 

 and ventral tracts is decidedly Caprimulgine. AVhether this indicates 

 a nearer approach to that hypothetical, lost, parent form is, to say the 

 least, doubtful. More probably Strix has varied from the Strigine 

 stem in the same way, though to a greater degree, perhaps, than Clior- 

 deiles has from the Caprimulgine. 



The conclusion, then, to Avhich this study of their i^terylography has 

 brought me is that the Caprimulgi are related to Striges, and not very 

 distantly either — probably a branch from the early j^art of the 

 Strigine stem. Dr. Sharpe, in his address at Budapest on '' Kecent 

 Attempts to Classify Birds," says that the idea that Caprimulgi and 

 Striges are nearly allied ''is now scouted," but he admits that the 

 nearest approach to the latter is in Steatornis. Garrod, in his very inter- 

 esting account of the latter genus,* concludes that it resembles Striges 

 much more than Caprimulgi, while Parker t considers its resemblance 

 to either group as being purely analogous, and so forming no connect- 

 ing link between the two. The weight of argument perhaps, of author- 

 ities certainly, is thus directly opposed to the conclusions to which 

 my observations had led me. It must be, therefore, as above stated, 

 that a conclusion based on one set of facts only is eminently unreliable 

 and should be set aside if the other characters are all against it. How- 

 ever this may be, I can only say that a comparative study of the ptery- 

 lography of the two groups as represented in North America most cer- 

 tainly shows some surprising similarities. Perhaps, however, it is only 

 an extraordinary case of what may be called ''analogous variation." 



Proc. Zool. Soc, London, 1873, p. 526. 

 0pp. cit., 1889, p. 101. 



