100 



S. IKEXO : 



tlie Table of C'oluui-s, p. 9(J). Tlie i'Vofispriug were composed as follows : 



Table II. 



Fj-generation (1016). 



Tims the actual miiubers of tlie three classes of the /''^-oflspriug agree ^^•ith 

 the expected cues, either on 9 : 3 : 4 or on 2 : 1 : 1 basis, though the agi'ee- 

 iiient is luuch closer iu the former tluiu in the latter case. The fiual decisiou 

 which alternative will l)e here realised, ^vould howe'S'er be possible oiily after 

 the examination of the T^a-oflspring ; thus if the segi'egatiou under discussion 

 Avill belong to the 2:1:1 type all yellows should be heterozygous, and all 

 oranges homozj-gons, whilst if it will belong to the 9:3:4 type some yellows 

 should be homozygous, and some oranges heteroz3-gous. Unfortunately on 

 account of the poor germination of seeds the jP,-generation contains very few 

 individuals, and the results are consequently rather imjierfect, because I was 

 able to examine the i^-oftspring of only 2 oranges, 3 whites and 3 3-ü1Io\\s 

 iu all of which generally vcr^- few seeds came to germination, viz. : — ■ 



1 orange has thrown almost exdusiveh" oranges (22 iu aU),' 



1 orange has segi-egated into 9 oranges and 2 wliites, 



3 whites have thrown whites only (19 in all), 



1 yellow has thrown only 1 yellow, 



1 yellow has segregated into 5 yello\\s tiud 1 white, 



1 yellow has segregated into ß oranges and 1 white. 



TTie fact that the i-atio of segi-egatiou in oiu- case should be 9 : 3 : 4 mav be 

 concluded from tlic presence of one heteroz^-gous orangt; just cited, liecausj iu 

 the 2:1:1 type no orange should be heteroz^'gous. This orange has imdergone 



1 Besides 22 oranges 1 magenta was founl among the progeny. A» disctisscd in Chapter 

 •■ M\itations, etc.", I (j). 121 flf) this 1 magenta is rcgnnleil to have lieen produced liy the reverse 

 mutation, and consequently the orange parent is consiilered to he of the composition WoOft'hh 

 (s. al.so Table VIII, No. 9 ; discussion, p. 125). 



