UTRICULAR STRUCTURES. 181 



chlorophyll-utricles, the smallest diameter is shorter than 

 it was at first ; in the thin, fibre -like red utricles both the 

 shorter diameters are very considerably diminished. 



The contents of the colour-utricles are at first always 

 homogeneous, green (fig. 12 a; 13 a\ red, yellow, or 

 blue. They may present different characters during the 

 life of the utricle. 



In the first place, the contents remain for the whole 

 period homogeneous and coloured, without alteration, or 

 at least without any that can be noticed. 



Or, secondly, one or more very minute granules become 

 visible in the contents, which are permanent in this stage, 

 and the nature of which cannot be made out, on account 

 of the small size. Probably they are starch- globules, as 

 would result from the following. In a third condition of 

 the contents, minute granules of the same kind appear, 

 but are somewhat larger, and appear as whitish globules, 

 coloured violet or blue by iodine (fig, 10, a, b, c ; fig. 11 ; 

 fig. 13; fig. 17). Mohl thinks that the violet tinge is 

 produced by the brown colouring of the surrounding 

 green contents. It appears certain to me, however, that 

 this is proper to the starch-globules. Even before the 

 treatment with iodine a difference is observable in the 

 latter, for they refract light more strongly and the others 

 more weakly. If they refract light in this manner, they 

 will acquire the blue colour, in spite of the surrounding 

 chlorophyll ; if the refraction is less, the colour is violet. 

 These starch-globules are very frequently solitary in an 

 utricle ; sometimes, however, from two to five are present. 

 They only occupy a small portion of the cavity of the 

 utricle. 



A third condition of the contents consists in the fol- 

 lowing : As before, minute granules first become visible 

 in the homogeneous, coloured contents. They increase 

 in size, and at last almost wholly fill the utricles, so that 

 merely a thin layer of colour remains investing the starch- 

 globule (fig. 12 i). If, as is generally the case here, the 

 membrane of the utricle is indistinct, it appears as though 



