504 
GENETICS: W, E. CASTLE 
secondly, the remaining observations are treated in utter disregard 
of the principle of 'interference' which has long been recognized by 
the authors and has never been repudiated by them. 
But we are not without other and independent evidence of the 
relations between the three genes, yellow, white and bifid. Each of 
these has known linkage relations with other genes of the system. 
Pertinent facts from Table 65 of Morgan and Bridges are reproduced 
herewith in Table 2. They show that yellow is farther removed from 
vermilion, miniature and rudimentary than is either white or bifid, 
since the cross-over values are in every case higher. This supports 
Morgan's view that yellow is at the distal end of the system. If 
white and bifid lie in succession below yellow, white should be farther 
removed than bifid from genes lower in the system. Data are given 
TABLE 2 
CROSS-OVER 
VALUES 
CASES 
34.5 
13,271 
34.3 
21,686 
42.9 
2,563 
31.1 
2,724 
30.6 
219 
42.7 
899 
42.5 
306 
30.5 
27,962 
33.2 
110,701 
42.4 
6,461 
45.7 
3,664 
Yellow- vermilion 
Yellow-miniature 
Yellow-rudimentary. 
Bifid-vermilion 
Bifid-miniature 
Bifid-rudimentary. . . 
Bifid-forked 
White-vermilion. . . . 
White-miniature. . . . 
White-rudimentary. , 
White-forked 
for the distances to four such genes. In two cases (vermilion and 
rudimentary) white is nearer than bifid, in two other cases (miniature 
and forked) bifid is nearer than white. The more reliable cases, based 
on larger numbers, are those which place white below bifid, minimal 
observations 2724 and 899 respectively. The numbers which place 
bifid below white are 306 and 219, too small to carry much weight. 
We may conclude that so far as evidence given by table 65, Morgan 
and Bridges, is concerned, bifid is certainly as remote as white and 
probably more remote than white from the lower parts of the linkage 
system. If so, bifid cannot lie in line with yellow and white, below 
them in the linkage chain, as maintained by Morgan and his associates. 
This conclusion agrees with the relations shown in my model, but 
not with those shown in their Diagram 1. 
