62 Acton. — On the formation of sugars 
saccharon is said to be almost completely insoluble in cold 
anhydrous alcohol. 
In the formation of glucose 1 from starch or cellulose, dex- 
trins are always formed simultaneously during the early stages 
of the reaction, whether the change is brought about by the 
agency of diastase or of artificial reagents, and the dextrins 
first formed (e. g. erythrodextrin) give a reddish colour with 
iodine. Now, as stated above, no substance showing the 
reactions of an erythrodextrin was detected at any stage in 
the cells, and I consider that the absence of such, especially at 
the time when the sugars can first be detected, is of some 
value as evidence that the glucose is not formed by hydrolysis 
of a carbohydrate. 
Although there is evidently a close connection between the 
groups of saccharons and glucoses, cane-sugar has not been 
artificially prepared by any reaction, and it does not seem 
possible to obtain it by any simple process from starch or 
cellulose. Maltose 2 , a carbohydrate having the formula 
C 12 H 22 O u ,- is easily obtained from starch, etc., but differs 
greatly in its properties from cane-sugar. 
The conclusions I should draw from these experiments as 
to the nature of the process of secretion of sugars in Narcissus 
and other plants having the kind of nectaries called septal 
glands, are : — 
1. That the first stage consists in a maximum formation of 
protoplasm containing a large amount of metaplasm, especially 
in the form of proteid granules, but not of starch-grains, 
mucilage, or any form of solid carbohydrate. 
2. That the sugars are probably derived from the decomposi- 
tion of this metaplasm, and constitute one of the products of 
the change. That both glucose and saccharon are formed 
simultaneously. 
3. That the excretion of the saccharine liquid 3 into the 
1 See O’Sullivan, Joum. Chem. Soc. xxix. 479, and xxx. 126 ; Brown and Heron, 
in Joum. Chem. Soc. xxv. 618. Musculus and Gruber, Comptes Rend. 86, 1459 ; 
Roscoe and Schorlemmer, loc. cit. ; Armstrong and Groves, loc. cit. 
2 See O’Sullivan, loc. cit. ; and Schultze in Ber. Deut. Chem. Ges. vii. 407. 
3 Compare Wilson in Unters. Bot. Institut, Tubingen, 1881 ; and Gardiner, loc. cit. 
