CHAP. V. BORAGO OFFICINALIS. 185 



on the opposite sides of four small pots, and treated as before. 

 But many of the plants were unhealthy, and their heights were so 

 unequal some on both sides being five times as tall as the others 

 that the averages deduced from the measurements in the pre- 

 ceding table are not in the least trustworthy. Nevertheless I 

 have felt bound to give them, as they are opposed to my general 

 conclusions. 



The seven self-fertilised plants from the crossed plants here 

 average 15 '73, and the seven self-fertilised from the self-fertilised 

 21 inches in height ; or as 100 to 133. Strictly analogous ex- 

 periments with Viola tricolor and Lathyrus odoratus gave a very 

 different result. 



XXIII. BOEAGINACE^:. BOEAGO OFFICINALIS. 



This plant is frequented by a greater number of bees than 

 almost any other one which I have observed. It is strongly 

 proterandrous (H. Muller, ' Befruchtung,' &c., p. 267), and the 

 flowers can hardly fail to be cross-fertilised ; but should this 

 not occur, they are capable of self-fertilisation to a limited 

 extent, as some pollen long remains within the anthers, and is 

 apt to fall on the mature stigma. In the year 1863 I covered 

 up a plant, and examined thirty-five flowers, of which only 

 twelve yielded any seeds ; whereas of thirty-five flowers on an 

 exposed plant growing close by, all with the exception of two 

 yielded seeds. The covered-up plant, however, produced alto- 

 gether twenty-five spontaneously self-fertilised seeds ; the exposed 

 plant producing fifty-five seeds, the product, no doubt, of cross- 

 fertilisation. 



In the year 1868 eighteen flowers on a protected plant were 

 crossed ^ith pollen from a distinct plant, but only seven of these 

 produced fruit ; and I suspect that I applied pollen to many of 

 the stigmas before they were mature. These fruits contained 

 on an average 2 seeds, with a maximum in one of three seeds. 

 Twenty-four spontaneously self-fertilised fruits were produced 

 by the same plant, and these contained on an average 1*2 seeds, 

 with a maximum of two in one fruit. So that the fruits from 

 the artificially crossed flowers yielded seeds compared with those 

 from the spontaneously self-fertilised flowers, in the ratio of 100 

 to 60. But the self-fertilised seeds, as often occurs when few 

 are produced, were heavier than the cr / >ssed seeds in ihe ratio oi 

 100 to 90 



