348 Mr. W. P. Pycraft on the 



primary and cubital series is almost twice the combined 

 length of the manus and foi'earm. The metacarpo-digital 

 remiges are not separated from the cubital series by a diastema, 

 as is usual, but are closely approximated, the space between 

 the two series being about equal to that dividing any two 

 " metacarpals/^ In other birds the " carpal diastema/' as I 

 propose to call it, is considerable — rarely less than the space 

 dividing any two of the cubital remiges, which are always 

 the more widely spaced. This crowding of the two series of 

 remiges seems to be correlated with low powers of flight. 



Of the cubital remiges it is interesting to remark that 

 usually one finds they are separated by fairly uniform spaces, 

 — decreasing proximally — whilst here the space dividing 

 Nos. 1 and 2 is less than the similar space separating each 

 of the remiges from 2-5, where the spaces appear to be rather 

 above what I should be inclined to fix as the normal width. 



Tec trices : — 



T. majores (woodcut, fig. l,p. 347, T.m.). — Usually the major 

 coverts on the dorsal surface of the manus are as perfectly 

 developed as those on the cubitus ; in Opisthocomus, how- 

 ever, it will be noticed that the feathers of this region are 

 much smaller than the cubital coverts. This is an arrange- 

 ment which occurs but seldom. I find it to obtain in such 

 forms, for instance, as Psophia, Dicholophus, and Musophaga. 

 Although possibly of no great significance, this is an inter- 

 esting fact. 



On the ventral surface this series is throughout reduced 

 to relatively small discontinuous feathers with a degenerate 

 hyporhachis (fig. 1, V.t.m.). 



T. medice (woodcut, fig. 1, p. 347, T.md.). — The cubital 

 coverts of this series on the dorsal surface increase in length 

 from without inwards, so that the most proximal feathers are 

 more than half as long as the major coverts underlying them. 



On the manus Nos. 1-4 are absent, as is often the case 

 with other birds ; the remainder of the series is fairly well 

 developed. 



On the ventral surface this series is represented only on 



