90 PERIOD IV. 



nothing 1 more than this, that Robert Brown recom- 

 mended the book to Charles Darwin, who found in it, 

 as he says, "an immense body of truth." 



In 1787 Sprengel had remarked that the bases of the 

 petals of Geranium silvaticum are beset with long- hairs. 

 Persuaded that no natural structure can be devoid of 

 meaning, Sprengel asked what purpose these hairs 

 might serve. A honey-gland in their midst suggested 

 that they might protect the honey by keeping off the rain, 

 which easily enters this shallow flower. Other honey- 

 secreting flowers were found to possess mechanisms 

 adapted to the same end. His first question suggested 

 a second : Why should flowers secrete honey? 



Malpighi had described the honey-glands of crown- 

 imperial (1672), and had seen that the honey must be 

 secreted by the petals, and not deposited from the 

 atmosphere, according to the notion then current. 

 Kolreuter (1761) had showed that insects may effect 

 the pollination of flowers. Linnaeus (1762) had given 

 the name of nectary to the honey-gland. He thought 

 that the honey served to moisten the ovary, though he 

 knew of staminate flowers furnished with nectaries. 

 He also threw out the alternative conjecture that the 

 honey is food for insects, which disperse the pollen by 

 their wings. Sprengel improved upon all his prede- 

 cessors, and made it clear that transference of pollen is 

 the main purpose of the honey in flowers. He was put 

 on the right track by the study of a forget-me-not 

 flower. Here he found the honey protected from rain 

 by the narrowness of the corolla-tube, whose entrance 

 was almost closed by internal protuberances. The 

 protuberances were distinguished by their yellow colour 

 from the sky-blue corolla, and this conspicuous coloura- 

 tion led Sprengel to infer that insects might be thereby 



