6o Flowers and their Unbidden Guests. 



any creeping animal must necessarily pass before it can 

 reacli the flowers. The variety presented by these 

 is very great. Sometimes it is the epidermis of the 

 bracts which itself directly secretes the viscid matter ; 

 the epidermis, therefore, acting just as it does on the 

 flowering axis of Oislus ladaniferus, Ledum palustre, 

 Lychnis viscaria ; sometimes, on the other hand, the 

 secretion is discharged from trichomes, the so-called 

 glandular hairs, which grow in countless number and 

 endless variety from the epiderm of the bracts, and 

 give the involucre its villous aspect and its stickiness. 

 Prom the vast number of forms which come under this 

 heading, I select a single one as an example ; and of 

 this I have given a drawing at Plate I. fig. 9. This 

 represents on an enlarged scale a small portion of 

 the capitulum of Grepis paludosa L., with three of the 

 marginal ligulate flowers, each with its anther-tube 

 and with its style protruding beyond this tube. The 

 style is divided into two branches, which have bent 

 themselves back in an arch, so as to expose the 

 stigmatic surfaces as freely as possible to any flying 

 insects. Such insects, if they would suck the nectar 

 which is to be found at the bottom of the corolla-tube, 

 can scarcely help brushing against the stigmas, placed 

 as these are ; and as the insects also visit flowers that 

 are in an earlier stage, in order to get pollen, their 

 visits will, as a rule, promote intercrossing, Crawluig 

 insects, on the other hand, would be able to get up 



