268 



THE SECOND-CHROMOSOME GROUP 



continued drop for the third and fourth broods (table 117). The only- 

 significant deviation from this rule is in the cavse of curved speck, which 

 rose slightly in the second broods (table 118). This probably means 

 that the normal fall was more than compensated for by the concomi- 

 tant change in the amount of double crossing-over in this particular 

 region. The coincidence values for the experiment show a very slight 

 drop with age, but here again the intervals are so long that conciseness 

 is lost and the real effect on coincidence obscured (table 119). 



Table 118. — Crossing-over values for successive broods of the 

 star-purple curved speck back-cross. 



Table 119. — Coincidence values for different regions of the 

 star purple curved speck back-cross. 



Table 120.— Pi, star 9 X dachs d"; B. C, Fi star 9 

 X dachs cf from stock. 



The star purple curved speck back-cross did not allow of a close 

 calculation of the locus of star because of the great distance between 

 star and purple, with little knowledge of the amount of coincidence 

 involved. For this reason it was sought to use dachs as the base of 

 reference for star, since dachs is about 23.5 units nearer star than 

 purple is. Streak was known to be even nearer, but the classification 



