58 THE ELEMENTARY STRUCTURE OF PLANTS. 



in internal or subterranean parts, most concealed from the light, 

 as in roots and tubers, the pith of stems, and seeds. Starch 

 consists of oval or rounded grains, usually somewhat irregular in 

 outline, and sometimes becoming polyhedral by mutual pressure, 

 as in rice. The size of the grains varies extremely in different 

 plants, and even in the same cell ; as in the potato, where the 

 larger grains measure from ^<ju to -^ of an inch in their larger 

 diameter, but the smallest only T V<y f an mcn - I Q wheat- flour 

 the larger grains are 1 ^ to ^^ of an inch in diameter. And the 

 largest starch-grains known are ^^ of an inch long. Indeed, 

 from their manner of growth, we 'might expect that their bulk 

 would be somewhat indefinite. The mode of their formation is 

 indicated by the peculiar markings, by which starch-grains may 

 almost always be recognized ; namely, by the dot or darker point 

 which is seen commonly at one end of the grain, and the fine 

 concentric lines drawn around it, which present the appearance 

 of a succession of irregular circles over the whole surface, in 

 whatever direction the grain is turned. These appearances are 

 best seen in starch from the potato, one of the most characteristic 

 forms and easiest to be examined, under a magnifying power of 

 from 250 to 500 diameters (Fig. 51). The chemical composition 

 of starch is exactly the same as that of cellulose (27) ; and the 

 grains are solid throughout, but their interior usually softer or more 

 gelatinous. The lines, therefore, it is evident, mark the concen- 

 tric layers, or hollowed scales, of different density, which are suc- 

 cessively deposited on an original nucleus. The dot (or hilum, as 

 it has been called) that indicates the position of the nucleus, be- 

 comes a concavity, from its not receiving a part of the successive 

 deposits, which are greatest on the opposite side, or very eccentric. 

 The grains lie loose in the cell, and are probably formed so ; 

 although it is thought by some that the nucleus or hilum was in 

 contact with the cell-wall, so that the increase by deposition must 

 necessarily have taken place on the other sides. On the whole, 

 there is reason to conclude that starch-grains are formed on nuclei 

 or cytoblasts, that is, on minute solidified portions of protoplasm, 

 like those from which cells primarily originate, by the deposition 

 of layer over layer of ternary assimilated matter (dextrine, &c.), 

 essentially like that which constitutes the secondary deposit that 

 thickens the cell-membrane (39). Their origin, therefore, would 

 be closely analogous to that of cells formed "directly from a cy- 



