THE ANGIOSPERMAE 1305 



are prickly and beset with protuberances, which suggests that they are 

 designed for insect polHnation and it is a fact that small flies do often settle 

 on Duck Weed and may effect pollination. Various types of water insects 

 such as the water skaters may also assist. Water snails, however, are fre- 

 quently found crawling among Lemna plants and have been frequently 

 observed to pollinate the flowers. This is particularly interesting because 

 these plants are protected by the presence of raphides from being eaten by 

 water snails. 



D. Pollination by Insects (Entomophily) 



The first observations which led to the realization that insects were 

 responsible for pollination were made by Koelreuter and published by him 

 in 1 76 1. This preliminary account was followed by two further papers 

 which appeared in 1764 and 1766. Koelreuter's first observations were 

 made upon the Fig and it was he who first appreciated that insects were 

 essential for the production of its seeds. In his later writings he records that 

 it came as a great surprise to him to discover that what was true of the Fig 

 was equally true of a number of other plants, and he mentions particularly 

 the Cucumber, the Sword Lilies (Iridaceae) and the Mallows. As his 

 observations multiplied he gained a clear conception of the principles of 

 cross-pollination and of the significance of dichogamy. Koelreuter's 

 observations, however, were soon surpassed by those of Sprengel, whose 

 investigations were carried out on a far wider scale, and the accounts which 

 he gave of his results raised the subject to an important branch of botanical 

 work. 



There are certain fundamental differences in the structure of the pollen 

 grains, which in general distinguish those which are destined to be distri- 

 buted by wind, water or insects. Wind-distributed pollen grains are gener- 

 ally small and smooth-walled, while the grains of water-distributed pollen 

 usually lack an extine. On the other hand, the pollen which will be distri- 

 buted by insects is far more variable in form. Generally it is adhesive, the 

 extine being beset with small spines, warts, pits and grooves, while in some 

 cases the grains are oily or are bound together with threads of a sticky sub- 

 stance which also assists in attaching the pollen to the stigma. Further, with 

 the exception of hydrophilous forms, the pollen of most Angiosperms is 

 immediately affected by water. In most instances, if water reaches mature 

 pollen grains, they are damaged beyond recovery and cease to be viable. For 

 this reason we find many ways in which plants contrive to protect the pollen 

 from the effects of rain and dew. It is impossible here to list all the methods 

 of protection which have been recorded, but mention may be made of a few 

 of the more typical. (See also p. 1154.) 



In many flowers the anthers are sheltered by a covering which may 

 either be formed by the flower itself or may be produced from some adjacent 

 structure. Among the former there are types in which the flower bends over 

 so that the anthers hang downwards and are protected by the calyx or 

 corolla. Alternatively either the flowers or the inflorescence may bend over 



