THE THBEE FOEIIS OP LTTHBUM SAIICAHIA. 185 



tained (excluding one capsule with 12 seed), on an average, 117-2 

 seed. 



Fifteen flowers fertilized by the longer stamens of the short- 

 styled form yielded 93 per cent, of capsules, which (excluding 

 four capsules with less than 20 seed) contained, on an average, 

 102-8 seed. 



Thirteen flowers fertilized by the shorter stamens of the long- 

 styled form yielded 54 per cent, of capsules, which capsules (ex- 

 cluding one with 19 seed) contained, on an average, 60-2 seed. 



Twelve flowers fertilized by own longer stamens j'iolded 25 per 

 cent, of capsules, which (excluding one with 9 seed) contained, on 

 an average, 77-5 seed. 



Twelve flowers fertUized by own shorter stamens yielded not a 

 single capsule. 



Considering the three Tables and this summary, we may safely 

 draw the following conclusions. First, that, as in structure so in 

 function, there are three females or female organs : this is mani- 

 fest ; for when all three receive the very same pollen, they are 

 acted on most difiierently. So conversely with the thirty-six sta- 

 mens, we know that they consist of three separate sets of a dozen 

 each, dilTering in various respects ; and in function the pollen of 

 these three sets when applied to one and the same stigma acts 

 most differently, as a glance at the Tables proves. But we shall 

 presently see that the action of the pollen of the whole dozen 

 longest and of the whole dozen shortest stamens is not identical. 

 Secondly, wo see that only the longest stamens fully fertilize 

 the longest pistil, the middle stamens the middle pistil, and the 

 shortest stamens the shortest pistil. And now we can compre- 

 hend the meaning of the almost exact correspondence in length 

 between the pistil of each form and the two half-dozen sets of 

 stamens borne by the two other forms ; for the stigma of each 

 form is thus rubbed against the same spot of the insect's body, 

 which becomes most charged with the proper pollen. In all tliree 

 forms, the female organ is but feebly, or not at aU, acted on by its 

 own two kinds of pollen. In my papers on the dimorphism of 

 Primula and Linum, I used the terms " heteromorphic " for the 

 fully fertile unions between the female element of the one form 

 and the male element of the other, and " homomorphic " for the 

 less fertile or quite sterile unions between the female and male 

 elements of the same form. The principle involved in these terms 

 holds good with Lythrum, but is insufficient ; for though in each 

 of the three forms the fertile unions are all heteromorphic, the ap- 



