22 Linkage Group I. 



consulting tables 2 and 3 in connection with the following consider- 

 ations, the data on each locus may be evaluated as they are taken up 

 for treatment. Table 3 gives the data for each gene in the order of 

 treatment, together with the experiment number and the number 

 of flies involved. The latter indicates the relative value of the experi- 

 ments in respect to numbers, and the experiment number permits 

 reference to table 2, which indicates the number of genes involved 

 in the experiment and other details that aid in evaluating it. 



Sepia. 



In previous papers yellow has been placed at the zero-point on the map; but sepia, 

 as shown by experiment 19 (table 2) occupies a position "above" that of yellow, and 

 thus becomes the terminal gene. Sepia and yellow give approximately 5.6 per cent 

 crossing-over, placing yellow at about 6. 



Yellow. 



Yellow is one of the most frequently used characters, and its locus has served as 

 a base in numerous experiments, as shoAVTi by table 3. Its relation to sepia has been 

 considered under sepia (experiment 19). This placed yellow at 5.6 units from sepia. 

 With crossveinless it averaged 18.7 per cent crossing-over, with vermilion 19.6 per 

 cent and with vesiculated 18.8 per cent. All of the latter three values should be 

 increased slightly by correction for double crossing-over. 



Frayed. 



Frayed was only used in three experiments before the stock was accidentally lost. 

 One of these (experiment 2) was with forked. It gave almost 48 per cent of crossing- 

 over, showing that the frayed locus is remote from that of forked. Another was 

 with yellow (experiment 1, Metz, 1918) and indicated very close linkage (1.3 per 

 cent crossing-over). The third was with vesiculated and magenta (experiment 14, 

 Metz, 1918) and indicated a locus about 18 units to the left of vesiculated. On the 

 basis of these, frayed is tentatively placed 1 unit below yellow. That it is close to 

 yellow is clear, but it is possible that the position should be above rather than below 

 that of yellow. 



Crossveinless. 



Weinstein (1920) has placed crossveinless at 17.6 units to the right of (or below) 

 yellow. Our data, summarized in table 3, give a value of 19.6 units for this "region," 

 based on 3,093 flies, without correction for double crossing-over. The value of 18.7 

 per cent used for calculating map length is based on the combined data (5,450 flies). 

 By adding this 18.7 per cent to the 5.6 per cent given by sepia and yellow, the value 

 24.3 per cent is obtained. Expressed in round numbers, this gives crossveinless a 

 locus at 24 on the map. 



Vermilion. 



Preliminary tests indicated that vermilion was very closely linked to crossveinless, 

 and experiments 20, 36, and 46, involving yellow, crossveinless, and vermilion, placed 

 vermilion slightly below crossveinless. The average of all data (table 3) gives a 

 cross-over value of about 0.47 per cent between the two. This would place vermilion 

 between crossveinless and vesiculated, although the sum of our crossveinless-ver- 

 milion and vermilion-vesiculated values does not agree very closely with the cross- 

 veinless-vesiculated value (1.2 per cent) obtained by Weinstein (1920). Our tests of 

 vermilion with vesiculated (table 3) give a value of approximately 2 units. 



The exact value depends upon how the data are treated. If all the flies are 

 counted the result is 2 per cent. But since vesiculated occasionally fails to appear 



