of Cellulose in Starch-grains. 2 to 



grain : in this case it would be necessary to assume that there 

 existed in the starch-grain, as in a glass globe whose tempera- 

 ture diminishes from the periphery to the centre, a dilating 

 tension in the radial direction, and a compressing tension in the 

 direction of the surfaces of the concentric laminae. It might be 

 urged in favour of this, that distinct reasons for the adoption of 

 this view exist in the behaviour of unaltered starch-granules 

 when dried or caused to swell up (as Nageli thinks it proba- 

 ble, and I believe justly, that in the fresh granules molecules of 

 water contained in them are interposed in greater number in the 

 tangential than the radial molecules of the substance) ; but we 

 cannot, at all events, conceive such a kind of tension to exist in 

 those granules in which the starch-substance has been extracted 

 from the outer lamina? by saliva, and in which, as above-men- 

 tioned, these outer laminae become torn, by cracks which run in 

 a radial direction, into polygonal fragments, since they cannot 

 contract over the still unexpanded inner laminae. Here it is 

 beyond doubt that there exists in these outer laminae an expan- 

 sive tension in a tangential direction, which rises to the point of 

 fracturing the laminae, while there is a compressing tension in 

 the radial direction, consequently the reverse of the above 

 assumption. In spite of this, these laminae freed from starch 

 act in the same manner as unaltered starch-grains. Indeed the 

 optical characters of the grains are not changed by the produc- 

 tion of any mechanical alteration in the starch-grains — or by 

 causing, through the action of alcohol or strong desiccation, the 

 formation of a central cavity and the retraction of the internal 

 layers towards the former periphery, combined with tearing in a 

 radial direction — or by causing a swelling up, by which the 

 outer layers then expand, especially in the direction parallel 

 to their surfaces. It therefore appears as though we had here 

 an example of those cases in which the optical phaenomena of a 

 body depend, not upon the density and relative distribution of 

 its substance, but upon the quality of the material, like the 

 conditions found by Greilich (Krystallogr. Opt. Untersuch. 

 1858, p. 226) in the comparison of isomorphous compounds of 

 potash and ammonia. In such cases we could not doubt that 

 the optical behaviour of the bodies depended on their intimate 

 molecular conditions. We must, doubtless, recognize this mole- 

 cular diversity — such as occurs, for example, in the various kinds 

 of sugar — in reference to its effects upon polarized light, first of 

 all as a physical property belonging to them ; but, since we have 

 no means of acquiring information regarding the peculiarities of 

 the molecules, and since these peculiarities run parallel with the 

 chemical composition of bodies, I believe I am perfectly justified 

 in assuming that the contrast of negative and positive colours 



