Compound Celluloses 103 



(i p.ct. NaOH). This treatment is carried out in duplicate, one 

 of the treated specimens being washed off with water, dilute 

 acid (acetic), and finally water, and dried, for the estimation of 

 the loss of weight in the alkaline treatment (a) The second 

 portion, after thorough washing from the alkaline treatment 

 and finally with distilled water, is squeezed so as to retain a 

 minimum of water, introduced into a glass bulb of extremely 

 thin walls, and sealed off. (b) The bulb is carefully introduced 

 into a bottle previously filled with chlorine gas, collected over 

 warm water and inverted with a glass plate placed on the mouth, 

 and in such a way that a minimum quantity of water is left 

 in the bottle. In any case a suitable quantity of coarsely 

 pounded glass should be introduced with the bulb, in order 

 to prevent the fibre being unduly wetted, which would retard the 

 absorption. The bottle is closed with an indiarubber cork, well 

 coated with paraffin, through which passes a bent tube. Through 

 this tube the bottle is brought into connection with any suitable 

 gas-measuring apparatus permitting accurate measurement of 

 the vacuum formed in the bottle as the reaction proceeds. 

 All parts of the apparatus being brought to a constant tempera- 

 ture, the bulb containing the fibre is broken by a blow against 

 the sides of the bottle. The chlorine is absorbed with rapidity, 

 and observations of the absorption are made from time to 

 time. If an apparatus such as a Lunge's nitrometer is used, 

 the apparatus is adjusted at its extreme mark, i.e. full of air. 

 As the volume of gas in the reaction flask shrinks, the liquid 

 levels in the measuring apparatus are adjusted in the usual 

 way. The reaction may be considered at an end when no 

 further absorption is noted during an interval of 10 minutes. 

 It is advisable to insert a stopcock between the reaction bottle 

 and the measuring apparatus, so that the latter may be cut off 

 after each observation of volume. 



In calculating from the observed numbers, it is only neces- 



