Trans. N. V. Ac. Sci. 40 ?««• 7. 



their own lower ones, closely approximate to, and often touching them ; 

 yet here no marginal irregularities are seen, except a mere roughening 

 of margins and midrib. Many of the orchids and liliaceae grow in 

 closely shaded places with leaves of other plants packed densely 

 around them, yet their leaves are entire. Only the palms, the high- 

 est of these endogens, bear compound leaves, but their pinnae are 

 entire. If Mr. Allen's hypothesis were true, we should certainly 

 expect to find marginal indentations on the leaves of most endogenous 

 plants having this habit of growth. The monocotyledons are plants 

 of a very old type, having appeared on the earth, geologically, long 

 before the angiosperms,* and hence there has been very much more 

 time for them to have undergone changes in leaf-forms, but these re- 

 main very much as they began. In their case there seems to be a 

 perpetuation of an ancient character dependent on a primitive plan 

 of parallel venation, which the imaginary "struggle" for carbonic 

 acid has failed to alter. 



The leaves of angiosperms hardly appear to have become more ser- 

 rated and lobed since their introduction into the earth's flora at the 

 beginning of the Cretaceous epoch. The earliest of them, as repre- 

 sented by specimens from the Dakota sandstones and New Jersey 

 plastic clays, are probably not less serrated nor lobed than the leaves 

 produced by plants of the same genera now living. That there has 

 been a change is of course possible, although it may justly be ob- 

 jected, that very little is known to us of that early angiospermous 

 flora compared with our knowledge of the present one. But we will 

 admit that a change in this direction may have taken place from nat- 

 ural causes, i.e., that leaves are now marginally modified to a greater 

 extent than they originally were, although the palseontological evi- 

 dence of this change is wanting. 



The object of the chorophyll-bearing parenchyma of the leaf is 

 primarily the absorption and decomposition of carbonic acid, and it 

 effects this in proportion to the amount of stomata-bearing surface ; 

 an entire-margined leaf accomplishes this more economically than a 

 lobed one with the same system of venation, for the surface is greater, 

 and the advantage of lobes, divisions, etc., under this view is difficult 

 to apprehend. 



Again, the well-known laws of the phenomena of diffusion of gases t 

 render it quite certain that the amount of carbonic acid in all parts 



* The oldest hitherto discovered plant of probable nionocotyledonous affinities is Potkocites 

 Gra>itoni, Paterson, perhaps allied to the Aracm, remains of which have been found in the 

 Carboniferous bituminous shale of Granton, Scotland, and described in Trans. Edinb. Bot. 

 Soc, vol. I. Undoubted endogens occur abundantly in the Trias and Jura. No true angio- 

 sperms are known lower than the basal Cretaceous strata. 



t See Ganot, Elementary Treatise on Physics, p. 135. 



