158 HAROLD H. PLOUGH 



region of the chromosome shows an unquestionable increase of 

 more than 50 per cent, no increase at all is registered in the test 

 between star and black. This can mean only that with such 

 long distances any increase in the actual amount of single cross- 

 ingover is compensated by a similar increase in the amount of 

 double crossingover, and thus no increase at all appears in the 

 percentage registered by the count. This fact emphasizes the 

 importance of working with 'short chromosome regions/ that is, 

 with factors between which there is only a low percentage of 

 crossingover, when testing for environmental effects on linkage. 

 The summaries of bottle counts which have been given so 

 far do not give any idea of the amount of variation between 

 individual bottles in any series, nor of the amount of overlapping 

 between the series. This can be gained in all cases by a compari- 

 son of the detailed summaries in the appendix. In view of the 

 large number of flies involved in table 7, it has seemed of value 

 to illustrate the percentage frequencies of the bottle counts 

 in the form of a curve. The results of this assembling may not 

 be as accurate as the weighed mean given in the summary be- 

 cause a large brood has no more weight than a small one, but they 

 tell more about the data given. In each case the standard 

 deviation is lower for the smaller chromosomal region, and the 

 fact bears out the statement already made that the shorter 

 region in every case is the more accurate. The increase in stand- 

 ard deviation of the heat series over the controls is not statisti- 

 cally significant. The fact that the means differ very little from 

 those given in table 7 shows that by either method the increase 

 due to heat is shown with equal clearness. 



Complete analysis of temperature effect 



After establishing the fact that extremes of heat and cold 

 acting on the females up to the time of hatching caused a decided 

 increase in the percentage of crossingover in the first ten day 

 broods, it became important to secure material for a table 

 which would show the steps by which this rise took place. To 

 do this series of tests were made, using ten different tempera- 

 tures ranging from the minimum to the maximum at which 



