100 TKE BOTANICAL MAGAZINE. [Vol. xxxn. No. 377. 



for the character in question but all the rest split and showed 

 to be heterozygous ones so that the}' make ratio as follows : 



I Het. 33 : Horn. 4 



V Het. 28: Horn. 3 



Totals Het. 61 : Horn. 7 



The ratio (61 : 7) is to be taken as 8 : 1, for in calculation 

 we get + 7.56 for" deviation from our expectation while standard 

 error is+8.25. 



Then let us turn to the nature of heterozygous ones in F 2 . 



Of the above mentioned 61 heterozygous individuals the 

 pedigree nos. 1, 3, 5, 6, 11, 13, 16, 17, 18, 19, 22, 27, 29, 31, 

 33, 34, and 51 from I lot, and nos. 1, 2, 3, 4, 5, 9, 11, 14, 15, 

 16, 17, 21, 22, 23, 49 and 51 from V lot split in the ratio 

 9 : 7 and all the rest in two lots split in the ratio 3:1 or 

 nearly so. The family-numbers of these two groups are 33 

 and 28 respectively so that they make ratio 1: 1 approximately. 



If we summarize the above numerical ratios of the hards 

 in F, (together with I and V) judged by the results in F 3 

 they are : — 



Fa hards 68 \H^'"Z , .'!.'.6l/ 33 inds - s P Ht into 9 : 7 



(28 inds. split into 3 : 1 

 Now let us see what relation we have between 33 and 28 

 individuals which split into 9 : 7 and 3 : 1 respectively. 

 In calculation as the ratio 1 : 1 we have : — 



Deviation from expectation + 2.5 



Standard error '. + 3.91 



The deviation is much smaller than standard error so that 

 we can think the two groups, i.e. (9 : 7)-group and (3 : 1)- 

 group, came out in equal numbers and furthermore when we 

 calculate the ratio based on the individual numbers for (9 : 7) 

 and for (3:1) groups respectively we can see the numbers are 

 all fit for expectation or in other words the deviation from 

 expectation is within its corresponding standard error thus : — 



1) Nos. 5 and 13 in I lot and 23 in V lot have the ratio 1:1 but this is 

 nearer 9 : 7 than 3:1, so they are included in the group of 9 : 7. 



