W. A. Davis, A. J. Daish and G. C. Sawyer 271 



pressed down and washed until the filtrate and washings have a volume 

 of nearly 2 htres. A Httle solid sodium carbonate is then added^ until 

 the lead is exactly precipitated, testing small portions so as to avoid 

 any considerable excess of sodium carbonate; the solution is then 

 diluted to 2000 cc. and a httle toluene added (1 cc.) to obviate bacterial 

 or fermentative change. We have found that a solution prepared in 

 this way can be left for several weeks without showing any change in 

 the proportion of sugars present; but care must be taken that the 

 solution is not left for any length of time, even a few hours, with any 

 excess of basic lead acetate or any considerable quantity of alkali, as both 

 of these substances rapidly destroy laemdose. 



The actual estimation of the sugars is described in a previous paper 

 [1913, p. 466]. 



Polarimetric Arrangements. 



With plant extracts such as those we have been studjang the purified 

 solution finally used for analysis is very dilute and therefore gives very 

 small angular readings in the polarimeter — generally less than 1° in a 

 200 mm. tube. For the results to be correct within 1 per cent., the 

 angular readings must therefore be accurate to within 0-01°. As slight 

 differences in temperature cause considerable alterations in the specific 

 rotatory power of laevulose and invert sugar, precautions must be 

 taken to ensure constancy of temperature during the observations. 

 All our readings have been taken exactly at 20-00°. By means of the 

 following simple thermostatic arrangement it is easy to maintain this 

 temperature to within ji^° for weeks together (Fig. 3). 



The temperature of the bath W , which consists of a large enamelled 

 iron vessel 16 ins. , 10 ins., is controlled by a Lowry toluene thermo- 

 regulator A ; the stirrer B is constructed from an old bicycle hub by 

 lengthening the spindle in both directions and attaching above a wooden 

 pulley 7 ins. in diameter, and below a four-bladed paddle. This stirrer 

 and the small Kohler centrifugal pump C are run from the same small 

 electric motor (gL h.p.) by a three-grooved pulley. The water-bath W 

 is kept covered by a tin-plate cover, with holes and grooves cut in it to 

 admit the neck of the thermo-regulator and other fixtures; a ther- 

 mometer (not shown in drawing) graduated in hundredths of a degree 

 also passes through this cover and enables the temperature to be read 

 accurately to ^^° C. The water of the bath, which is maintained at a 



^ When the content of sugais is too high to allow of the analysis being made directly 

 by Brown. Morris and Millar's method, .300 cc. of the 2 litres should be diluted to 500 cc. 

 Joum. of Agric. Sci. vn 19 



