210 THE STRUCTURE OF FLOWERS. 



one, of 97 seeds. In no legitimate union has so high an 

 average of 68 seeds been observed by me, or nearly so high 

 a maximum as 82 and 97." * 



I give these results of homostyled Auriculas and Chinese 

 Primroses as illustrating the principle so abundantly proved 

 amongst other plants — that as soon as they begin to retrace 

 their steps from a prevailing differentiated condition self- 

 fertilisation is rapidly resumed, and there follows a resumption 

 of a vastly increased rate of seed-making. They prove, too, 

 that however apparently stable these highly differentiated 

 states may normally be, various conditions of environment 

 can readily break them down ; thus, with cultivated plants, 

 usually so much stimulated, starvation is a potent cause. t 

 Linum perenne, as the above table shows, is particularly 

 barren when illegitimately fertilised, but a single branch on 

 a plant has been known to become homomorphic, and then 

 to set seed abundantly; this occurred with Mr. Meehan. 

 Warming found MenyantJies trifoliata to have become com- 

 pletely homostyled in Greenland. 



Trimorphic Flowers. — As a type of heterostylism where 

 a species adopts three f.orms, L. Salicaria may be taken. 

 Briefly summarizing Mr. Darwin's elaborate experiments on 



* Forms of Flowers, pp. 218-221. 



t It is not only true with heterostyled plants, but the rule applies 

 generally to highly cultivated flowers, that degeneracy from a floral point 

 of view is correlated vrith enhanced powers of self -fertilisation. Thus 

 a professional cultivator of Cyclamens is in the habit of keeping a stock 

 of "worthless" weedy-looking plants, for the express ptu^pose of raising 

 seedlings, as they are so much more prolific than the true florists' types. 

 Having obtained them, he then crosses them, and brings them up to the 

 standard required. Indeed, the fact is well known to all cultivators, that 

 the poorer the plant may be, from the florists' point of view, the better 

 seed-bearer is it; and that continually crossed and "perfect" flowers 

 are proportionally impotent or tend to become so, when a tendency to 

 become petaliferous often affects the essential organs. 



