224 OBSERVATIONS ON THE CHANGES OF 



rations of a naturalist, who was an e je witness of the facts he describes, 

 in preference to the mere conjectures of another. 



Although it has been more my intention in this paper to record the 

 facts which I have been able to collect on this subject, than to frame a 

 new theory, or endeavour to restore an exploded one ; yet it has often 

 occurred to me, that the advocates of the two opposite theories have 

 run into extremes, in consequence of the adherence of each, with too 

 much pertinacity, to a preconceived opinion. Hence Montagu, having 

 ascertained that some birds are subject to two moultings in a year, 

 conceived that this was the only mode by which the colours of feathers 

 were changed, asserting that "he had no conception of the feathers 

 themselves changing colour." On the other hand, Dr Flemming, and 

 those of his school, including the Rev. Mr Whitear and Mr Ord (the 

 latter advanced a step beyond all the rest, in stating as his opinion that 

 "all birds moult but once a year"), having observed that the colours 

 in feathers change in the early stage of their growth, came to the con- 

 clusion that the same process must always be carried on in the old 

 feathers, and that a fresh moult being unnecessary, would consequently 

 not take place. 



The observations which I have made thus far, have led me to con- 

 ceive that the truth lies between these two extremes. Although my 

 investigations were not extended to as great a number of species as I 

 could have wished, or repeated through a succession of seasons, it has, 

 notwithstanding, been very apparent to my mind, that young feathers 

 frequently change colour, particularly in autumn ; but that in those 

 cases where there is a semi-annual change of colours, there is in all, or 

 nearly all the species, a semi-annual moult. That a much greater 

 number of species change their feathers twice a year than is usually 

 supposed, and that our inquiries must be directed to individual species, 

 rather than to genera, since in the same genus one species is subject to 

 a double moult, whilst the others moult but once a year. 



4. Having now seen in what manner nature performs her operations 

 in effecting the changes in the plumage of birds, I proceed to inquire 

 whether the same laws may be applied to those variations in colour to 

 which the hair of quadrupeds is subject. The vascular bulbs in which 

 the roots of the hairs are inserted, bear an analogy to the bulbs which 



