80 REPORT — 1859. 



leave Cellulose undissolved, and thus serve as a means of its preparation, there is in 

 this no evidence that it is absolutely insoluble, and that the whole of it existing in the 

 substance so treated, has remained undissolved. The recently discovered action of 

 an ammoniacal solution of oxide of copper proves that Cellulose exists in different 

 states of aggregation, or induration, in some of which it dissolves readily in this solu- 

 tion, whilst in others it appears to be quite insoluble in it. May not, therefore, Cel- 

 lulose in these different states possess different degrees of solubility when treated with 

 acids and alkalies ? Again, there is no proof that the Cellulose which remains undis- 

 solved after the limited action of these agents, will do so when the action is continued 

 for a considerable length of time. Indeed, the custom of late years, of designating 

 the Cellulose in Food-stuffs as — young and old — soluble and insoluble — digestible and 

 indigestible — seems to indicate a general opinion that the term Cellulose, as hitherto 

 employed, either includes several distinct bodies, or one that occurs in different states. 



The importance of this subject led me to attempt to solve some of the questions 

 involved, by means of direct experiment. 



A sheet of Swedish filtering paper was divided into five parts, which were respectively 

 treated as follows : — 



No. 1— digested, for half an hour, at 160° to 180° Fahr., with a mixture of 1 vol. 

 oil of vitriol, and 2 vols, water. 



No. 2 — boiled, for half an hour, with very dilute acid. 



No. 3 — boiled, for half an hour, with an alkali-solution containing 1 percent, alkali. 



No. 4 — after treatment with acid as No. 1, boiled with alkali as No. 3. 



No. 5 — after treatment with acid as No. 2, boiled with alkali as No. 3. 



The result was, a loss upon Nos. 1, 2, and 3, of from 1 to 2 per cent. This, how- 

 ever, is not more than might be due to the unavoidable loss in decantation. 



Nos. 4 and 5, on the other hand, suffered a loss of from 8 to 12 per cent. A part 

 of this was doubtless due to the same cause as that operating in experiments 1, 2, and 

 3 ; but I cannot imagine that the loss I witnessed of this kind can account for the 

 whole of that which occurred in experiments 3 and 4. In fact, I cannot doubt 

 that a considerable portion of the loss in the latter cases, was owing to a change 

 effected by the acid on a part of the Cellulose, by which it was rendered soluble in the 

 alkali. 



Besides this evidence against the assumed stability of Cellulose itself, the complete 

 separation from it of the matters associated with it in vegetable products, is by no means 

 so easy as it is generally described to be. The Cellulose I obtained by following the 

 principal methods that have been proposed, was in no case entirely free from nitrogen. 

 How far it may be possible to obtain it pure I shall not now consider. But, I may 

 state, that even when the action of the solvents was carried far beyond that to which 

 the substance is subjected in the process of analysis, some nitrogen still remained with 

 the Cellulose. 



It would appear, therefore, that determinations of Cellulose made by alternate treat- 

 ment with acid and. alkali, can only be accepted as approximate. 



The question arises — how far the approximate results obtained by the different 

 methods that have been employed, are likely to be comparable with one another ? To 

 determine this, a series of experiments was arranged with a view to show the effect 

 of each of the solvents usually employed, and also, separately, that of each of the 

 attendant conditions of concentration, temperature, and time of action. 



In all cases finely ground hay was the material operated upon ; and, in each set of 

 experiments, the effect of the variation of one single condition was made the point of 

 inquiry. It should, too, be observed, that in none of the experiments did the degree 

 in which the several conditions were employed, reach the limit that is found to be 

 recommended as proper for the determination of Cellulose. 



The influence of variation in the strength of the acid, was the first point I endea- 

 voured to determine. To this end, equal quantities of the finely-ground, and dried 

 hay, were boiled, for a quarter of an hour each, with equal volumes of dilute sulphuric 

 acid, the strength of which was different in each experiment. The undissolved residue 

 was in each case well-washed, and then boiled, for a quarter of an hour, with a dilute 

 solution of potash, containing I per cent, of the hydrated alkali (KaO, HO). The 

 emaining insoluble matter was thoroughly washed, dried, and weighed. The time 

 action of both the acid and the alkali, the temperature t and the strength of the 



