C)02 Miissus. J. LKWis BoxnoTi': axd f. w. smalley ox 



not readily admit of a scientifie interpretation unless the birds 

 themselves can be seen ; frequently the same colour (from a scien- 

 tific point of view) may be known by different names according to 

 the particular breed of pigeons that may be under discussion. 



Darwin's experiments related chiefly to "reversion," i.e. the re- 

 appearance of the blue colour when distantly related breeds of 

 pio-eons were crossed, and although our work has not been on the 

 same lines, the study of the blue colour should, none the less, bear 

 out the results ari-ived at years ago. We are not unmindful of 

 this, and for that very purpose hope to mate up several pairs to 

 test Darwin's conclusions. 



Lastly there is Mr. Staples-Browne's recent paper, in which he 

 has attempted to repeat Darwin's experiments. This paper we 

 have been through very carefully, and find that in the main it 

 afyrees with our results, but one or two little difHculties have to be 

 met, such as the occurrence of a Dun in Exp. 27, the absence of 

 Blacks in Exps. 9 & 11, the large proportions of white in 

 Exps. 16-23 and 30, and the occurrence of white feathers on a 

 homozygous blue. 



"VVehave no doubt that further woi-k will clear up these slight 

 difficulties, which do not, however, greatly afJect the main result ; 

 in fact a possible explanation of some of them has already suggested 

 itself to us, but the discussion of these is best deferred until the 

 results of our matings on the same lines as Darwin and Staples- 

 Browne have been obtained. 



The majority of the birds used have been highly bred Dragoons, 

 but in a very few cases a Homer cross has been introduced. 



The characters dealt with in this paper are : — 

 (i) Colour-patterns, i. e. Chequering, Grizzle, and JMealy. 



(ii) Colours, i. e. Blue and Silver, Avith White and Red iv:i those 

 cases where it is connected with Grizzle and Mealy. 



Before detailing the experiments, however, it is necessary to 

 have a clear understanding of the terms used. 



(i) Chequering. This is a pattern chiefly confined to the wiug- 

 coverts, in which each chequered feather has a light coloured 

 V-shaped patch at its distal end, the apex of the V being- 

 nearest the base of the feather (see PI. XXVI. fig. 1). 

 The general appearance of a good chequered bird is shown 

 in PI. XXIIl. fig. 3, but it must be understood that 

 scientifically and in the experiments we have carried out, 

 a bird has been considered as chequered when it showed 

 the characteristic markings on its wings, 

 (ii) Grizzling. A grizzled feather is one in which the barbs are 

 partially white and pai^tially coloured. This pattern is not 

 restricted to any paxticular part of a bird, but grizzled 

 feathers may be found in any feathei- tract including the 

 remiges and rectrices. 



On a grizzled bird (/. e. a bird with grizzled feathers) 

 w-hole coloui-ed featheis are generally found as well as 

 feathers splashed with white (Pi. XXIY. figs. 1 &, 2; 



