————ncitidliaiioe 
862 ON EPIDERMAL CHLOROPHYLL. 
Assimilation appears to be more vigorous in the grains of the 
stomatal guard-cells than in that of the other epidermal elements, 
and this is the case even where, in the latter, the amount of assi- 
ted material is greatest, as in Bryonia divica, the Solanaceae, 
and Mirabilis longiflora: nevertheless no traces of starch were 
detected in the guard-cells of Lupinus hirsutus, Borrago officinalis, 
and sometimes of Anagallis arvensis, and there were but faint hints 
of it in Lactuca sativa. 
have also observed that in very poor light assimilation does 
not take place in epidermal grains: this suggested the thought that 
perhaps the negative results had been obtained from plants examined 
early in the morning and after some days of dull weather, but re- 
un the starch reaction is 
dependent upon the depth of the grains’ colouring. Tt must be 
as to grades of colour is very difficult. Still the fact that in one 
with apparently quite as deeply-tinted chlorophyll, no marks of 
assimilation are visible, does seem to teach that assimilation is 
of t ; » £. rubrum, and Godetia rubicunda, have -in 
place of ordin nnin a substance taking wi e or 
| purple colour; it gives the tannin reaction with iron salts, but 
plant’s epidermis be mounted in weak iodine, a beautiful blue 
colour declares itself in all the cells. Weak iodine should always 
be preferred in testing for this body, which Kraus nevertheless 
thinks to be tannin, because, in a feebly alkaline fluid, a weak 
solution of iodine gives with tannin colours similar to those produced 
in the epidermal cells.* Whether this idea be well-founded it is 
difficult to say ; an alternative view is that the substance in question, 
although not tannin, is a closely related member of the tannin series. 
* Dufour (Bull. Soc. Vaud. d. Sc. Nat. 1886; abstracted in Bot. Ztg. 1886, 
p. 869) has found this “ soluble starch "which he regards as a carbohydrate— 
1300 examined. 
only in about 20 types out of 
