:i70 The t'arholiiid rates oj tlic Maiujuhl Leaf 



2-0 grms. of dry matter are weighed, to witliin a few milligrams, so that 

 the error involved in this method is considerably less than the probable 

 error of sampling. The apparatus shown in Fig. 2 consists of a small 

 rojoirf-bottomed flask with a light, ground-in glass stopper which carries 

 two side tubes as shown, one of which, A, passes to within an inch of 

 the bottom, whilst the other, B, only just enters the stopper. The two 

 tubes can be closed with glass plugs as shown. The flask and stoppers 

 are first weighed alone and then the 20 cc. of extract is evaporated in 

 the flask as far as this is possible in the water-bath; the stopper is 

 then inserted and the flask placed in a copper Meyer bath, filled with 

 toluene, the end of the tube A being connected with the vacuum pump, 

 a flask of phosphorus pentoxide being interposed to absorb moisture. 

 An ordinary Meyer bath serves to heat two of these flasks at the same 

 time, the copper lid of the bath being replaced by an asbestos cover 

 with two holes cut in it to take the nocks of the flasks. In all cases the 

 dry matter was heated 18 hours in vacuo at 110°, the weight being then 

 practically constant. Unless the drying be carried out in vacuo in the 

 manner described it is impossible to obtain anything like constant 

 results. 



The apparatus shown in Fig. 2 serves also to estimate the tnoislurc 

 in the leaf matter left after extracting the sugars (see p. 268) prior to 

 the estimation of starch. 



Estimation of sugars. The 440 cc. remaining of the original 500 cc. 

 are transferred to a large flask, diluted with about 300 cc. of water and 

 exactly the necessary quantity of basic lead acetate^ added to precipitate 

 the whole of the tannins, amino-acids, etc. present ; care should be taken 

 to avoid any considerable e.xcess of the basic lead, which must be added 

 in small quantities at a time, until on filtering a small portion of the 

 solution and testing it no further precipitate is produced. Working in 

 this way it is possible to avoid having more than 1 or 2 cc. of the lead 

 acetate solution in excess at the end of the precipitation. The quantity 

 of lead acetate solution used- with different samples varies widely, 

 according to the nature of the leaf, the time of year, etc. ; when 

 1 kilogram of mangold leaf is used, the 440 cc. of the sugar solution 

 requires from 200-300 cc. After the precipitation is complete the 

 solution is filtered on a large Buchnor funnel ((3 ins.) aiul the precipitate 



' We show in a sep-iratc paper that, by using basic load acetate in the inaiuier we 

 pi escribe, no loss or destruction of the sugars is to be feared. 



' We use the ordinary solution of basic lead acetate as employed in general sugar 

 analysis (sp. gr. 1'25, sec Allen's Commercial Organic Atwli/sis, vol. i. p. 308). 



