384 The Physiology of Plants Book iii 



always invests them. The increase of the thickness of the 

 grain at any point is proportional to the thickness of the 

 enveloping plastid at that point. Meyer was thus the first 

 to show that the starch grain in the leucoplast, as well as 

 in the chloroplast, always arises and remains in its interior. 

 Meyer's work was supplemented in the next year by his 

 pupil Salter. 



Schimper was thus the founder of the views at present 

 held as to the deposition of starch in the reservoirs of 

 storage. The relation between the plastid and the starch 

 grain, or the probable process of deposition was in- 

 vestigated with some minuteness. The plastid itself, as 

 described by its discoverer, has a fairly simple structure ; 

 Schimper held that these bodies always arise from ante- 

 cedent plastids by division, while other workers have spoken 

 of them as being differentiated from the protoplasm in the 

 cell. Schimper described them as oval or rod-like pieces 

 of protoplasm, frequently containing a crystalloid of protein 

 nature, which he held to be nutritive for the plastid, and 

 to disappear as the starch grain grows. Meyer confirmed 

 him on these points. 



The manner of formation of the starch grain by the 

 plastid was attributed by Strasburger in 1882 to a process 

 similar to that which he had observed in the case of the 

 thickening of the cell wall, where micro-somata existing in 

 the protoplasm are gradually deposited over its surface, 

 the layer so formed slowly ceasing to give the reactions 

 of protoplasm and showing instead those of cellulose. He 

 held that the starch grain is built up in the same way, each 

 successive layer showing a transition from protoplasm to 

 starch. This mode of formation was also advanced by 

 Noll in 1887 in slightly different words. He held that the 

 layer of protoplasm lining the cell wall becomes charged 

 with carbohydrates, and as these increase the protein mole- 

 cules gradually retreat from the layer, so that ultimately 



