298 CALVIN B. BRIDGES 



of crossing over between black and purple shows an initial high 

 value (8 per cent) which during the first nine days falls rapidly 

 at first and then more slowly to a low value (5 per cent), which 

 is maintained with little change to about the sixteenth day. A 

 sharp rise then sets in which reaches its maximum (8 per cent) 

 at about the twenty-first day. The succeeding fall is again 

 slow, reaching its minimum (3.5 per cent) at about the thirtieth 

 day. Beyond this point the curve again rises slightly, but the 

 data were too few to be significant beyond about the twenty-fifth 

 day. While there was some variation in the amount and 

 rapidity of these changes in the various individual curves, all 

 showed the same typical rhythm, which must be the expression 

 of fundamental physiological changes in the development of the 

 female. It seems possible and probable that these successive 

 falls and rises are not effects of a single continuously varying 

 physiological process, but are rather to be explained as separate 

 phenomena caused by the lapse of certain conditions and the 

 subsequent onset of new causes. These changes may therefore 

 be really discontinuous and the rhythmic curve only a succes- 

 sion of independent but overlapping variations. 



The most interesting feature of the age variation is the bear- 

 ing it has on the problem of double crossing over and the 

 underlying problems of the nature of crossing over. The more 

 consideration that has been given to this problem of double 

 crossing over in relation to chromosome and to map distances, 

 the more involved it has appeared, so that no evidence upon 

 these points can be neglected. The special value of such cases 

 as that of age variation is that they enable one to compare two 

 different conditions, but with the elimination of one important 

 variable; for, the actual chromosome distance between two 

 given loci is maintained constant, so that any variations that 

 occur in map distance, coincidence, etc., must be due to varia- 

 tions in one or more of the other factors. This relation was 

 discussed in connection with the original two-brood black purple 

 curved experiment (Bridges, '15), and it was pointed out that 

 the rise in coincidence concomitant with the fall in crossing over 

 meant that the internode length had changed — that the two 



