48 



types of comb, viz., s.s. singles, r.s. 

 roses, p.s. peas, and finally r.p. combs, 

 the latter being a type not before seen 

 in the course of the work.^ It is char- 

 acteristic of their comb that it is wide 

 like a rose, though much less papillose 

 and with no posterior point or "pike"; 

 but the most singular feature of it is 

 the presence of small bristles or iiiiinite 

 feathers on the posterior third. Often 

 these feathers form a narrow band 

 crossing the comb from side to side in 

 front of the posterior third. Sometimes, 

 however, there is no definite band, but 

 the feathers are either generally dis- 

 tributed over the back or on the pos- 

 terior sides of the comb. No such 

 bristles or feathers ever occur on 

 singles, roses, or peas. Their existence 

 raises a morphological problem of 

 some interest, but this cannot be 

 treated as yet. In front of the feathered 

 part of the comb is often a well- 

 marked transverse groove. On ap- 

 proaching maturity the part anterior 

 to the groove generally widens out 

 (especially in males) and becomes cor- 

 rugated, the posterior part remaining 

 narrow and flat. 



This peculiar comb is the structure 

 known as the "walnut" comb, proper 

 to Malays and to no other breed of 

 fowls. As stated, it was here first 

 formed indirectly by crossing r.s. X 

 p.s. and it has since been made directly 

 by breeding both pure r. and extracted 

 F2 r.r. with pure p. Sometimes it is 

 distinctly larger than in pure Malays, 

 but the artificial walnut comb is often 

 indistinguishable from the pure Malay 

 type. 



The results from the "artificial" wal- 

 nut combs were as follows. Three such 

 r.p. $, 's were bred through the season 



3 The features of the 3 forms of dominant 

 were not satisfactorily distinguished in the 

 earlier chicks from Experiment 124. They 

 arc therefore given collectively as 68 r., r.p., 

 p. and 24 s. 



BATESON AND PUNNETT 



of 1903 with various s.c. birds (Ex- 

 periments 163-170, 174, 176, 176a-b), 

 and gave— 



r.p. r. p. s. 



139 142 112 141 + 2 doubtful. 

 Simple 



expectation 133 133 133 133 



Four "artificial" r.p. $ 's with simi- 

 lar 5's gave (Experiments 162, 172 

 and 173, 175)- 



r.p. r. p. s. 



95 26 38 16 



Simple expectation! 99 33 33 11 



We have, therefore, in the first case 

 a fair approach to equality, and in the 

 second, figures not very far removed 

 from 9 : 3 : 3 : 1 . It is, therefore, clear 

 that the artificial r.p. birds were giving 

 off approximately equal numbers of 

 gametes, r.p., r., p., s. The individual 

 results show some irregularities, espe- 

 cially 176b, which gave 10, 17, 7, 8, 

 but the average result is unmistakable. 



These facts show two points of in- 

 terest, first, that some gametes are 

 bearing r.p. unsegregated; secondly, 

 that s. is evidently present though it is 

 presumed that in the parental gametes 

 of the r.s. and p.s. birds the s. elements 

 had been eliminated by segregation. 



At first sight it seemed that the pres- 

 ence of s. indicated a failure of segre- 

 gation. Further experiment, however, 

 has shown that r.p. birds raised from 

 pure r. and pure p. birds are in exactly 

 the same condition as those raised from 

 r.s. X P-s. This latter observation, 

 which will be described with the work 

 of 1904, proves that s. gametes may in 

 some way be formed by resolution of 

 the product of the cross r. X p- 



The next step was to investigate the 

 properties of the true Malay walnut, 

 or natural r.p. comb, and especially to 

 determine whether it could be re- 



