connected together that a wide and deep funnel is formed by 

 their union. This funnel is so constructed as to facilitate the 

 entrance of insects, and yet effectually to prevent then* escape ; 

 one part of it is smooth, aud this slippery path leads down to a 

 tank of water at its hottom, where the insects are di-owned, 

 while a set of hairs and spines pointing downwards effectually 

 bar the way for ascent and escape. Other forms of traps 

 occur- in other plants. In Pinguicula the outer margin of the 

 leaf folds backwards, and in Utricularia curious pockets with 

 valvidar openings are distributed over the floating roots to 

 catch water-fleas, &c. Besides these traps, aU known 

 insectivorous plants have a cimous set of glands or organs 

 for the elaboration and absorption of their insect-food. These 

 glands are well seen in the leaves of Pinguicula. They are 

 cu-cular in form, flat or dome-topped, composed of cellular 

 tissue, divided into four or eight compartments, and contain- 

 ing protoplasm, with some small granules of chlorophyll. 

 Their function seems to be to absorb and appropriate such 

 parts of the entrapped uisects as are fit for the food of the 

 plant to which they belong. 



In conclusion, I may mention that I consider the vast 

 world of animalcular Ufe as another source from which plants, 

 especially such as are furnished with hairs, down, or other 

 form of tangle, may derive a supply of nitrogenous food. The 

 following quotation from Mr. SaviUe Kent's ' Manual of the 

 Infusoria' (part 1, p. 140) now in course of publication first 

 suggested this idea to my mind, and will, I thiak, convince 

 you that this source of supply, though constituted as it is ol 

 the minutest of organised bodies, is by no means an in- 

 significant one. "On October 10th, 1879, a day of intense 

 fog, the author gathered grass saturated with dew from the 

 Kegent's Park Gardens, the Eegent's Park, and the lawn of 

 the Zoological Gardens, and submitted it to microscopical 

 examination without the addition of any supplementai-y 

 liquid-medium. In every di-op of water examined, squeezed 

 from the grass or obtained by its simple apphcation to the 

 glass slide, animalcules in their most active condition were 

 found to be literally swarming, the material derived from each 



