474 THE SURVIVAL OF THE UNLIKE. [xXX. 



This is a section of a Trophy. If we were to ex- 

 amine a hundred specimens of this variety we should 

 find no two alike in shape and number of cells, and, 

 consequently, in shape and size of fruit. Moreover, 

 we should find the variations to be very great. Now, 

 fruits in wild nature possess a definite number of 

 cells, and of definite shape. The Trophy, then, is a 

 monster ; it is unnatural. To find a fruit nearer the 

 original wild type, we must find one more constant in 

 its character. We examine critically every large - 

 fruited sort, and we find each one monstrous in re- 

 gard to form and number of cells, but some are less 

 so than others. The least monstrous are always those 

 with the fewest cells. The fewest -celled fruits in our 

 garden, then, must be nearest the original type. Fig. 

 2 represents a sectional view of a normal Criterion. 

 The cells are three, incomplete. The fruit. Fig. 3, 

 is oblong, mostly regular. The smallest, most regu- 

 lar specimens of this variety are incompletely two- 

 celled. Fig. 4. On the other hand, abnormal 

 specimens of this variety are many -celled, as shown 

 by the partially -lobed fruit in Fig. 5. Occasionally 

 the tendency to monstrosity extends to the flowers, 

 and a twin is the result, Fig. 6. The Criterion pre- 

 sents nearly the whole record of development within 

 itself. Its regular, small, normal, two -celled fruits 

 approach the original type. Figs. 5 and 6 attest an 

 excessive influence of cultivation. All the fruits 

 here represented grew upon the same plant. The 

 Criterion must be compared with the pear-shaped 

 and egg-shaped sorts. Fig. 7 represents one of the 

 Pear tomatoes. It is almost uniformly two -celled, 

 or, in its larger form, the King Humbert, it becomes 



