844 
Journal of Agricultural Research voi. xxiv, No. w 
Tabi,:^ XVIII .—Families in which only rogues appeared. Genetic constitution of F., 
parents was XX YY, XX Yy, or XXyy 
Culture No. 
Stipule 
ratio of F2 
parent. 
Rogues. 
0.410. 
55 
49 
50 
25 
23 
52 
30 
18 
0.427. 
2.25 
0 . 4?4 . 
0.442. 
Ck.AAA .. . . ^ 
2.35 
0 . 44 t;. 
.. . . . 
. 
0.46^. 
0*476. 
16 
29 
15 
21 
0.480. 
0.481. 
0.48^. 
18 
0.48*;. 
26 
2.45 
24 
17 
20 
0.488. 
0 . 40 «;. 
0.406. 
13 
42 
39 
14 
34 
35 
‘ry''.. 
0.408. 
0.400 . 
o.i;o2. 
0.2421. 
0.2422. 
Culture No. 
1 Stiptile 
ratio of F2 
1 parent. 
i 
Rogues. 
1 
0.242 . 
: 42 
! 38 
32 
61 
0.2426. 
2.55 
0.2421;. 
0.24^6. 
0.24^7. 
44 
41 
! 34 
36 
i 
^ 27 
20 
33 
28 
0.2440. 
0.2418. 
0.2446. 
0.24^0. 
0 . 241;4 . 
. 
o.24<;6. 
0 . 24*;7 . 
2.65 
31 
19 
51 
17 
23 
18 
34 
44 
15 
17 
0.2460. 
1 
0.2461;. 
0.2460. 
0.2471 . 
.... 
0.2472 . 
0.2470 . 
2-75 
0.248^. 
0.2486. 
2.85 
0.2487. 
.. 
Tabi,^ XIX .—Ratio of XX : XoF plants among 132 F2 segregates of Gradus.rogue X 
Mummy, as determined from analysis of the F3 generation, based on data in Table 
XVII. The expectation is two segregating and one nonsegregating F^ families 
Nonsegre¬ 
gating, a 
XX. & 
Segregat- 
ing.a Xx' 6 
Total. 
Observed. 
46 
44 
2 
3-5 
•5 
86 
88 
2 
3-5 
•5 
132 
132 
Calculated. 
Difference (D). 
Standard error (S. E.). 
D/S. E.. 
o Character of F3 family. & Constitution of F2 parent. 
The total number of rogues (2,097) in the cultures containing few 
nonrogues is obviously far in excess of the total number of nonrogues 
(308) that would be expected if normal Mendelian segregation occurred 
in the Fg plants heterozygous for factors X and x' (Tables XVI, XVII, 
XVIII). In the families producing many nonrogues and few rogues 
the situation is reversed, the cultures containing in all 70 rogues and 311 
nomogues, plainly not in accord with the theoretical, monohybrid 3 : i 
ratio shown in the same tables. 
Disregarding the two t)^es of Fg families and considering all the 
segregating families as a whole, the observed numbers are: 2,167 rogues 
(XX and Xx') and 619 nonrogues (x'x'). The numbers actually found 
in the two groups show an excess of 78 individuals in the rogue class, a 
deviation of nearly 4 times the probable error of the difference between 
the observed and calculated number of variates (Table XX). 
