894 MB. C. BAUWIN ON THE ILLEGITIMATE OFFSPRI>'0 



size of the pollen-grains, and often in colour. The stamens are 

 graduated in length, go that one of the two sets in two of the 

 forms is equal in length to the pistil in the third form. For in- 

 stance, in the long-styled form the pistil equals in length the 

 longer set of stamens in the mid-styled and short-styled forms. 

 In all three forms the union is fully fertile and legitimate only 

 when the pistil is impregnated with pollen from the stamens 

 which equal it in length. Thus the long-styled form can be 

 legitimately fertilized only by the longer stamens of the mid- 

 styled or short-styled form ; it can be illegitimately fertilized 

 by its own two sets of stamens, and by the shorter stamens of 

 both the mid-styled and short-styled forms ; so that the long- 

 styled form can be fertilized legitimately in two ways and illegi- 

 timately in four ways. The same holds good with the mid- 

 styled and short-styled forms ; hence with triraorphic species 

 eighteen unions are possible, of which six are legitimate, and pro- 

 duce legitimate offspring, and twelve are illegitimate and produce 

 illegitimate offspring. 



I will give the results of my experiments on the illegitimate 

 offspring of various dimorphic and trimorphic plants in full detail, 

 partly because the observations are extremely troublesome, and 

 will not probably soon be repeated (thus I have been compelled to 

 count under the microscope above 20,000 seeds of Lytlirtf/m Sali- 

 carta), but chiefly because much light is thus indirectly thrown on 

 the important subject of hybridism. 



Lythrum Salicaeia. 



I will commence with this trimorphic species. Of the twelve 

 illegitimate unions, two were completely barren ; and I succeeded 

 in raising seedlings from only six, or doubtfully from seven, of the 

 remaining ten illegitimate unions. The experiments are arranged 

 in classes according to the parentage of the plants. In each case 

 I give the average number of seeds per capsule, generally taken 

 from ten capsules, which, according to my experience, is a nearly 

 sufficient number. I give also in each case the maximum number 

 of seeds in any one capsule ; and this is a useful point of compa- 

 rison with the normal standard — that is, with the number of seeds 

 produced by legitimate plants when legitimately fertilized. I 

 give likewise in each case the minimum number. When the 

 maximum and minimum diflfer greatly, and no remark is made on 

 the subject, it maybe understood that the extremes are so closely 

 connected by intermediate figures that the average is fair. Larga 



