HOUWINK’S EXPER. CONC. THE ORIGIN OF SOME DOMESTIC ANIMALS 9 
tral pair of rectrices, which are much longer than the others, soft in 
texture und curved, resembling the upper tail-coverts. These are about 
twice as long as the second pair, and four times the length of the outer 
rectrices. The first primary is considerably shorter than the tenth, the 
fifth being the longest. The tarsi are longer than the middle toe and 
armed with well-developed spurs”. 
„The feathers of the rump are long and lanceolate, and the hackles 
of the neck are of the same character, or if truncate have a specialized 
curve in the vane near the tip, and show iridescent colouring”’. 
„All four species will cross with one another, and the hybrids are mo- 
re or less fertile among themselves. The moult is typically Phasanine, 
that of the tail being from the outer rectrices inward. In the females the 
comb is rudimentary, while the wattles, specialized hackles, central rec- 
trices and spurs are lacking. Thus we see that the secondary sexual cha- 
racters are the comb, wattles, hackles, central tailfeathers and spurs. 
The only one of these characters which may be taken, to distin- 
guish the genus is the comb. The compressed tail is found in Lophura, 
Acomus, Lobiophasis and others; the lateral wattles and the hackles 
in Chrysolophus, and the median wattle in Tragopan”. 
„The four species fall into two quite distinct but unequal divisions, 
good subgenera they might be called; first: gallus, lafayetti and sonne- 
yati, and second: varius. These have been considered as two genera 
(Gallus and Creagrius) by a fewwriters e.g. GHIGI 1905. Varius posses- 
ses the peculiarities of a smooth-edged comb, a median throat wattle, 
truncated neck hackles and an extra pair of rectrices. Taking the group 
as a whole, however, the hiatus between the four species and the near- 
est related genera seems much greater than between the two groups 
themselves. So I choose to keep them together. And here comes the 
question of logicality ;whether by doing this I have not been somewhat 
illogical in comparison with other generic divisions. This is of not the 
slightest moment to me. No two genera of the Phasianidae or any other 
group of organisms, as now isolated by time and space on earth, are se- 
parated from each other by exactly the same intervals of character dis- 
tinction. Classification, we all admit, is merely the make-shift, inci- 
dent upon, and made necessary or indeed possible by, our ignorance of 
intervening forms. Hence relative clarity of interrelations is its sole 
aim. In this instance the genus Gallus, considered as embracing all four 
forms, expresses much more exactly the homogeneity of the quartet as 
