STRUCTURE.] ELEMENTARY MEMBRANE. 9 



vegetable kingdom do not exhibit such a composition. Link 

 never could find the spiral structure described by Meyen in 

 the aerial roots of what he calls a Stelis. It is, however, very 

 remarkable, as Link observes, " that many portions of plants 

 have a tendency to split spirally ; this, however, only takes 

 place in the thicker parts ; for instance, in old porous vessels, 

 and even in bark, as in that of the birch tree. We need not, 

 therefore, go back with Mohl to a molecular theory, which is 

 better left to the ' philosophers ' of nature." (See p. 13.) 



Membrane, as true cellulose, may be regarded as being in 

 the beginning a gelatinous precipitate from the organic 

 mucus of vegetation. Like all such precipitates, it must be 

 understood to be a collection of globular molecules adhering 

 by their points of contact ; and hence its permeability. The 

 history of organic precipitates has been studied by both 

 Harting and Link, of whom the latter thus states the result 

 of his observations : ' ' All precipitates, when analysed imme_ 

 diately after their formation, exhibit globules ; these globules 

 unite themselves to larger ones (being therefore fluid, like 

 globules of quicksilver) ; and these united globules or drops, 

 subsequently only (and that frequently under our own eyes) 

 change themselves into crystals. If M. Harting did not 

 observe this, it was owing to his not having examined the 

 precipitates speedily enough. The globules sometimes form 

 flat surfaces, sometimes they are gelatinous. All fluid sub- 

 stances exhibit a commencement of solidity on their surface 

 for we attribute fluidity to a substance, if the parts can be 

 displaced from each other by the slightest application of 

 force ; and this can only be done, when the attracting and 

 repelling powers of the homogeneous parts neutralise each 

 other, which cannot be the case on the surface of fluid sub- 

 stances, where the parts are unequally drawn in different 

 directions. This solidity increases with the surface, and a 

 thin stratum of fluidity is consequently in itself solid. The 

 degree of solidity certainly depends on the degree of attrac- 

 tion among the parts, which, as is well known, is different 

 also in fluid substances, as exhibited by quicksilver and water. 

 Nothing, therefore, is required for the production of a 



