388 The Physiology of Plants BOOK in 



attention to many points of resemblance between starch 

 grains and sphere crystals in general, showing that swelling 

 does not change the optical properties of the grains, and 

 thus proving that Naegeli was wrong in attributing their 

 double refraction to internal tensions. Schimper held 

 that their ultimate constitution is one of crystalline fibres 

 arranged across the laminae and resembling protein crystal- 

 loids in their properties. 



Strasburger agreed that the laminae may be regarded as 

 composed of elongated radially arranged elements, but he 

 held that they are not necessarily crystalline. 



In 1887 Mikosch said that the grains of the potato are 

 made up of radially placed rodlets, which can be made 

 to break up into a granular mass. He held that the rods 

 consist of a ground substance in which the granules are 

 contained. 



Meyer's views of 1895 accorded with those of Schimper. 

 He developed the idea in greater detail ; besides showing 

 that their properties agree with a sphaero-crystal constitu- 

 tion, he was able to render the radial trichitic structure 

 visible by suitable means, frequently by inducing a slight 

 swelling of the grain. 



As has been already mentioned, although Schimper's 

 discovery of the plastids was the basis of most of the 

 recent work and may be regarded as well-established, there 

 are cases on record which constitute exceptions to it. 



Strasburger, in 1882, found that in Marsilia and in the 

 medullary rays of Pinus, starch grains arise in cavities of 

 the cytoplasm. Salter offered as an explanation of the obser- 

 vation the suggestion that the grain seen was the staining 

 kern, and the ' vacuole ' the colourless layer (rand) sur- 

 rounding it. In 1887 Belzung claimed to have seen them 

 arise in the meshes of reticulated protoplasm. In 1893 the 

 present writer found them in pollen grains as shed from the 

 anthers of various plants, and noted a great formation of 



