Letters, Extracts, Notices, ^c. 473 



rience, is correct. Mr. Seebohm's favourite theory of in- 

 terbreeding is all very well in some cases, but it certainly 

 does not apply in this instance, as there is only one form of 

 hen bird. 



In our paper, at p. 336, line 13, there is, unfortunately, a 

 clerical error, which we corrected in ' The Ibis ' for October 

 1884, p. 466, and to which we again call attention. Under 

 the head 6th stage, for " We consider stages 6 and 7 to be 

 of about the same period of the bird's life,'^ read '' We con- 

 sider stages 5 and 6/' &c. 



Mr. Seebohm misunderstands us in saying " It is alleged 

 that the brown of the underparts changes first to white, then 

 to black, kc," as will be seen by referring to the paragraph 

 below the 9th stage, in which it is clearly stated that the 

 brown or blackish brown changes to pure white, the lower 

 parts never change, so far as our knowledge goes, from white 

 to black afterwards. In some instances, however, the white 

 head is gained before the white belly (6th stage), while in 

 others the white belly is assumed before the head changes as 

 described in the 5th stage, the result being that at this stage 

 birds which have not shed their abdominal feathers are still 

 black below, others which have commenced the moult are 

 black with traces of white, and those which have completed 

 it are white. Stages 5 and 6 are therefore very similar, the 

 diflfereuce between them being that in the 5th stage the bird 

 has assumed the white lower parts, and in the 6th it has not 

 completed the moult in respect to those parts. The 6th 

 stage is not very common, and we only obtained a few ex- 

 amples in that plumage, and most of those showed traces of 

 white below, as previously stated, showing the gradual change 

 from black to white. We consider stages 5 and 6 to be two 

 phases of the one stage, and they might perhaps have been 

 advantageously united. Mr. Seebohm considers that it would 

 be as logical to make the variations nineteen as nine in num- 

 ber ; but for the purpose of explaining our theory we found 

 the nine stages given sufficient. 



In conclusion we may remark that, notwithstanding Mr. 

 Seebohm's decision, we are still of opinion that our theory is 



