IN THE TROPICS. 
381 
seeds to white approaches 3 : 1 ; and further that among 
the coloured seeds the ratio of those with purple spots to 
those without also approaches 3 : 1. The total proportion, 
therefore, would appear to be 9 : 3 : 4. (A ratio also found 
by Tschermak (63) to exist in several similar cases.) 
At first sight this does not appear to be a Mendelian ratio, 
but written beneath the proportion for a Mendelian di¬ 
hybrid case the agreement is clear ; thus : 
9 : 3 : 3 : 1 
9 : 3 : 4 
Now suppose the French sugar pea to include two allelo¬ 
morphs : A giving a testa coloured greenish-grey like the 
common field pea ; and B giving purple spots when A is 
present, but ineffective when A is absent For the white- 
coated pea we may write ab (absence of the greenish-grey 
colour and of purple spots). 
AB x ab gives— 
Fj ABab 
(Gametes : AB ; Ab ; aB ; ab.) 
F 2 AB + 2ABa -f- 
2ABb + 4ABab + 
Ab + 2Aab + 
Ba 
2Bab 
ab 
This obviously gives the required proportion. In F 3 we 
should expect to find— 
. 1 constant purple ••• ... AB 
2 purples giving purple and field-coloured 2ABb 
2 purples giving purple and white ... ABa 
4 purples giving purple, white and field-coloured 4ABab 
1 constant field coloured ... Ab 
2 field-coloured giving f. coloured and white 2Aab 
4 constant white ••• ...Ba + 2Bab + ab 
Twenty to forty seeds were sown from each of the 
following F 2 plants :— 
29 with purple-spotted testas ... (p) 
8 with field-coloured testas .. (f) 
7 with colourless testas ... (w) 
