18 EXPERIMENTS IN THE HYBRIDISING OF INDIAN COTTONS. 
No. L appears to have been a pure yellow, and No, 19a 
pure white-flowered plant. Nos. 2 to 12 gave between them, out 
of a total of 676, 166 whites or 24°2% of recessives. On the 
Mendelian theory, the number should have been 169—a very 
close approximation. 
TaBLeE X. 
Lhe fifth hybiid generation. 
| ri | i 
No. in | TOTALS, 
No table Grandparent. Parent. Yellow. |, White. 
IV. Y; W. 
| 
——) —___+ 2 cet = ii 
1 2 | Yellow | Yellow 7 13 * Al 
2 8 Do. \- Doz He 13 : 
3 9 Do Do. 9 
4 | 15 Do | Do. ae 28 a soe <a. 
a 4 18 Do. cam OS * 53 | ct 116 0 
6 | 2 Do mol eo: 129 54 < ; 
7 8 Do Do. 108 44 
8 9 Do Do. 26 6 
9.4) 285 Do Do. 177 ys 
10 16 Do. %: Do. ae 83 27 o Py 
11 18 Do. Wee besa! Die): a 90 26 613 215 
12 2 Do. White d 1 SI fs aa 
13 6 Do. Do. 6 
14 8 Do. Do. 6 
15 9 Do. Do. 5 
16 18 Do. Do. i 2 
17 19 Do. Do. 1 86 3 193 
18 13 White Do. 5 41 z 
19 Do. Do. 1 31 4 72 
The difference between yellow and white flowered plants is 
again plainly shown. The first five plants were pure yellows, 
and gave not a single white in over a hundred plants. The next 
six were hybrid yellows, and from them out of 828 plants 215 or 
26 “{ were whites, a very close approximation to the expected 
ratio of 25 %. Extracted whites gave of nearly 200 plants, 
only 1:5 % yellows. The last two plants were in small beds 
among yellows to vicinism with which the yellows among them 
are probably due. 
THE SEED. 
Cotton seeds differ in shape, in size, and in colour —7.e., 
after removal of the “ floss” or “lint” the surface may be 
included, the proportion of white to yellow or pale—33 to 101- would be almost exactly the 
expected Mendelian ratio of 1 to 3, 
