20 JOURX.IL OF THE EOYAL HORTICULTUEAL SOCIETY, 



The half should have violet-red flowers (Aa), Classes 1, 3' 



„ „ ,, white flowers (a) „ 2, 4 



.„ „ „ a long axis (Bb) „ 1, 2 



„ „ „ a short axis (b) „ 3, 4 



From 45 fertilisations of the second year 187 seeds resulted, of which- 

 only 166 reached the flowering stage in the third year. Among these the^ 

 separate classes appeared in the numbers following : — 



Class. Colour of flower. Stem. 



1 ^let-red long 47 times 



2 white long 40 „ 



3 violet-red short 38 „ 



4 white short 41 



There consequently appeared — 



The violet-red flower colour (Aa) in 85 plants. 

 „ white „ (a) in 81 „ 



„ long stem (Bb) in 87 „ 



„ short (b) in 79 „ 



The theory adduced is therefore satisfactorily confirmed in this trial also. 

 For the characters of form of pod, colour of pod, and position of flowers 

 experiments were also made on a small scale, and results obtained in 

 perfect agreement. All combinations which were possible through the 

 union of the differentiating characters duly appeared, and in nearly 

 equal numbers. 



Experimentally, therefore, the theory is justified that pea hybrids 

 form egg and pollen cells ivhich, in their constitution, represent in equal 

 numbers all constant forms ivhich result from the combination of the 

 characters ivhen conjoined by fertilisation. 



The difference of the forms among the progeny of the hybrids, as well 

 as the relative ratio of numbers in which they are observed, find a 

 sufficient explanation in the principle above deduced. The simplest case 

 is afforded by the developmental series of each pair of differentiating 

 characters. This series is expressed by the formula A-f2Aa-fa, in 

 which A and a signify the forms with constant differentiating characters, 

 and Aa the hybrid form of both. It includes in three different classes 

 four individuals. In the formation of these, pollen and egg cells of the 

 form A and a take part on the average equally in the fertilisation ; 

 hence each form twice, since four individuals are formed. There parti- 

 cipate consequently in the fertilisation — 



The pollen cells A -f A -f a + a 

 The egg cells A + A -f a + a. 



It remains, therefore, purely a matter of chance w^hich of the two 

 sorts of pollen will become united with each separate egg cell. Accord- 

 ing, however, to the law of probability, it will always happen, on the 

 average of many cases, that each pollen form A and a w^ill unite equally 

 often with each egg cell form A and a, consequently one of the two 

 pollen cells A in the fertilisation will meet with the egg cell A and the 



