THE FACTORS OP ORGANIC EVOLUTION. 61 



the structure of the species when injured. We cannot, 

 however, ignore the immediate influence of the medium, on 

 recalling the facts above named, or on remembering the 

 further fact that an inflamed surface of skin, when not 

 sheltered from the air, will throw out a film of coagulable 

 lymph. But that the direct action of the medium is a chief 

 factor we are clearly shown by another case. Accident or 

 disease occasionally causes permanent eversion, or protru 

 sion, of mucous membrane. After a period of irritability, 

 great at first but decreasing as the change advances, this 

 membrane assumes the general character of ordinary skin. 

 Nor is this all : its microscopic structure changes. Where 

 it is a mucous membrane of the kind covered by cylinder^ 

 epithelium, the cylinders gradually shorten, becoming finally 

 flat, and there results a squamous epithelium : there is a 

 near approach in minute composition to epidermis. Here a 

 tendency towards completion of the type cannot be alleged ; 

 for there is, contrariwise, divergence from the type. The 

 effect of the medium is so great that, in a short time, it 

 overcomes the inherited proclivity and produces a struc 

 ture of opposite kind to the normal one. 



With but little break we come here upon a significant 

 analogy, parallel to an analogy already described. As 

 was pointed out, an inorganic body that is modifiable by 

 its medium, acquires, after a time, an outer coat which 

 has already undergone such change as surrounding agencies 

 can effect; has a contained mass which is as yet unchanged, 

 because unreached; and has a surface between the two 

 where change is going on a region of activity. And we 

 saw that alike in the vegetal cell and the animal cell there 

 exist analogous distributions : of course with the difference 

 that the innermost part is not inert. Now we have to note 

 that in those aggregates of cells constituting the Metaphyta 

 and Metazoa, analogous distributions also exist. In plants 

 they are of course not to be looked for in leaves and other 

 deciduous portions, but only in portions of long duration 



