410 REVIEWS. 



grain, which also tends to become enlarged. This addition 

 of water, however, occurs more easily in a tangential direc- 

 tion, and each layer is more or less prevented from enlarg- 

 ing by the other closely adhering layers ; thus, as Naegeli 

 found, each lamella suffers a. positive tension with respect to 

 the next inner one, and is in negative tension towards the 

 outer ones. Hence also follows that the cut surface of a 

 halved grain becomes concave, and that the nucleus may 

 become hollow from the tugging exerted by the layers as 

 they increase their surface. The assumption of water can 

 only be great, however, in those layers which pass com- 

 pletely round the grain ; hence, in Phajus starch those 

 layers which only go partly around the whole grain show no 

 regular increase of watery contents as we pass from the 

 younger to the older parts. 



In certain cases the simultaneous action of several causes 

 produce complications into which we cannot here enter. 

 The radial slits, &c., so often found in the completely 

 formed granule, are co-ordinated with the differences in 

 cohesion in different directions. On drying, for example, 

 most water is abstracted from the internal parts, and the 

 cohesion is weakest in the tangential direction; hence we 

 find the resulting fissures are radially arranged, and are 

 wider towards the centre of the grain. So, too, the adhe- 

 sion of the lamellas being very strong, pressure and artificial 

 treatment causes no separation of the layers, only the for- 

 mation of radial fissures. The close correspondence be- 

 tween the structure of a starch grain and of a thickened cell- 

 wall, induces Strasburger to believe that the lamellae in 

 both cases must result from the union of " microsome " 

 series and sheets ; that the cohesion between the rows of 

 "microsomes" of any one sheet is weaker than that between 

 the sheets or lamellae. In Pijius wood-cells, for example, 

 the union between two series of '* microsome " rows may be 

 weaker or stronger — as was seen in cases where the sheets 

 broke up into spirals — and so on. 



In transferring the hypothesis from the cell- wall to the 

 starch granules, and vice versa, it must of course be remem- 

 bered that in the one case the lamellae are plastered on, so to 

 speak, from the inside, while in the other they succeed one 

 another from without. In an ordinary cell the straining 

 from within is balanced more or less by the pressure and 

 elasticity of the external layers ; pauses in growth enable 

 the environment to produce changes resulting in the forma- 

 tion of Grenzhdutchen, and this is most evident at the ex- 

 ternal surface, where a cuticularised layer is formed outside. 



