IN THE SECOND CHROMOSOME. 329 



other data known. But the other data for Cm C IIr show con- 

 clusively that b and p r give very little crossing-over (0.2 per cent), 

 while either with v g , c, or s p , gives about 1.1 per cent; and v g and c 

 give only 0.1 or 0.2 per cent with s p . That is, v g and c are on the 

 same side of b and p r . And the extensive data for b p r c show that 

 the sequence is b p r c. Therefore the one individual that suggested 

 the sequence p r b v g must have been a double cross-over. 



(2) In the case of C IIr only three cross-overs between c and s p 

 were obtained. Of these, two were also cross-overs between b and c, 

 while one was not. These data alone would indicate the sequence 

 as 6 Sp c, instead of the usual 6 c s p . No great significance can be 

 attached to the difference between 2 flies and 1 fly among a total of 

 1,615. In any case, the data suggest a very high coincidence. More 

 data of the same sort will be necessary before this exceptional case 

 can appear significant. 1 



(3) In the case of -79 -, only 3 cross-overs were observed 



^11 r 



between b and p r . All of these were also cross-overs between p r and 

 s p . 2 If the coincidence in this case is 100, approximately the value 

 usual for b p r c, then nearly half of the b p r cross-overs should be also 

 p r s p cross-overs. Therefore the fact that all 3 were such doubles 

 need not cause surprise; even though, taken alone, it would indicate 

 the sequence as p r b s p . 



The three exceptional cases are, then, of no great significance, ex- 

 cept as indicating rather high coincidence. There are a large number 

 of cases in which the evidence is much clearer and in which the sequence 

 is certainly the same as that usually found. 



HOW DO Cm AND C IIr PRODUCE THEIR EFFECTS? 



The question of the mechanism whereby the cross-over genes pro- 

 duce their effects is not yet satisfactorily answered. Cytological 

 examination might conceivably furnish the solution, but has not yet 

 been seriously attempted. A study of coincidence might give a clue, 

 but is difficult to make, because of the very small percentages that are 

 concerned. 



In the case of C n r and Cm it is to be noted that two like chromo- 

 somes cross over freely, while two unlike ones do not. 3 While this is 

 only a restatement of the facts, it at least offers an attractive opening 

 for speculation as to the nature of the case. 



J In a culture derived from Cm, n experiments discussed above, a female that was apparently 



C n f 

 of the constitution r (without Cm, it) has recently been tested. One a v daughter was 



J)f C 8p 



produced. If this record represents what it appears to, the count becomes 2 double cross-overs 

 against 2 single cross-overs. 



2 Two of them were also recorded as cross-overs between S' and 6 ; but this is probably incorrect, 

 as was pointed out above (p. ). Cm 



3 So far as the evidence goes, this is also true for Cn j, but 7; is unknown. 



V//I ( 



