ANATOMV Ol I'LANTtJ. 215 



In the Malvacea; and the Compositae, its form is tlie same, 

 (Tab. IV. Fig. 15.) It appears, in these instances, to con- 

 sist of regular spherical bodies, with bristly openings. In 

 the Liliacea3 and the Geraniums, it consists of oval-shaped 

 bodies, surrounded by elastic rings or hoops, (Tab. II. Fig. 

 24.) In the Onagra? we find obtuse triangular bodies con- 

 nected by slimy threads. In the Proteaceae the pollen is cy- 

 lindrical, and somewhat bent; (Bauer's lUustr. Nov. Holl. 

 Tab. III. Fig. H.) The two first forms seem to be very 

 common. But the pollen varies its shape when it is put into 

 water or oil. In the former it becomes inflated, takes a more 

 or less perfectly globular form, and gives out, often with great 

 elasticity, its contents, like a small cloud, which do€s not mix 

 with the water. In mild oil it remains pretty much un- 

 changed, only it becomes by degrees surrounded by a dark 

 coloured ring, which by and by passes into the oil. In spirit 

 of wine it contracts pmverfully, and takes frequently, when it 

 is treated with nitric acid, an obtuse triangular form, which 

 originates in the contraction of the three rings. In nitric acid 

 it gives out its contents in the shape of rays, which do not 

 unite with the fluid. 



386. 

 There are some remarkable variations from these forms of 

 the antherae and pollen, most of which we have already noticed, 

 (107.) There is, however, still one variety in the structure of 

 the pollen in some of the Naiadae, particularly in Chara and 

 Zostera, to which we must attend. In these instances, the 

 antherae seem to contain nothing but strings of conferva?, 

 which do not unite with the water. These are as distinctly 

 inclosed in particular reservoirs, as in tlio Fuci ; and in some 

 Ferns they stand single beside the germen. 



337. 



So many varieties, which might easily bo nuiiti plied, in the 



structure of the pollen, naturally lead us to a difference of 



chemical contents. Hitherto, however, experiments of tlii> 



kind have only been made with the pollen of common plants, 



