W. A. Davis, A. J. Daish and CI. C. Sawyer 



L>69 



chlorophyll or fatty matter is present it is necessary to wash out the 

 flask with a little hot alcohol or toluene ; in this way there is no difficulty 

 in transferring to the measuring flask every trace of soluble matter, 

 whether sugars or leaf fats. The solution is then diluted to 500 cc. at 

 1.5° and is usually fairly homogeneous ; when, however, toluene has been 

 added the mixture should be thoroughly shaken so as to form a tine 

 emulsion immediately before withdrawing each of the samples for the 

 dry matter estimations. Working in this way the results are usually 

 quite concordant, the differences seldom exceeding 0-2 per cent, on the 

 weight of dry matter. 



WS^ 



Fig. 2 Flask used to tstiraate dry matter in the alcoholic extract. 



For the dry mailer estimations, 20 cc. of the 500 cc. are transferred 

 by means of a pipette to the small flask shown in Fig. 2, evaporated 

 to dryness on a water-bath and finally dried in vacuo at 110°; three 

 samples of 20 cc. should be withdrawn, two being used for the actual 

 estimations and the third kept as a reserve in case the duphcates do not 

 agree sufficiently closely. Usually the duphcates agree, when say 1-5 to 



roUdorij fowern and portions of 50 cc. for inversion with citric acid and with invertase. After 

 inversion these portions are neutralised, diluted to 100 cc. at 15° and 50 cc. of the solutions 

 so obtained are used to measiu-e the increase of reducing power or change of rotatory 

 power caused by inversion. This 50 cc. corresponds with the 25 cc. used for the direct 

 measurements. 50 cc. of the 500 cc. are used to estimate the pentoses, and 50 cc. portions 

 are fermented with the special maltase-free yeasts to estimate maltose, controls being 

 also made with baker's veast. 



