1877.] 



Sugar in relation to the Blood. 



317 



doubtful nature. The result being given by the balance, there is no un- 

 certainty belonging to it, as may be the case to a slight extent where the 

 gradual fading of colour has to be watched and a decision formed with 

 regard to the attainment of the precise point. If the reduction of the 

 oxide of copper can be safely turned to account for effecting the quanti- 

 tative determination of sugar (and analysts are agreed that it can), it is 

 by such process that the most trustworthy information is supplied. The 

 closeness attainable in the results of counterpart analyses shows that it 

 there is not only susceptible of being carried out with great precision, but 

 affords strong evidence of its reliability. Compared with this process, 

 the results yielded by that of Bernard present the greatest discordancy. 

 The figures given by Bernard's method are almost invariably too high, but 

 there is no uniformity in the difference presented. The variance, indeed, 

 shows no intelligible relation, and suggests that there is something fun- 

 damentally wrong in taking decolorization, without precipitation of the 

 suboxide, as a means of estimating the amount of sugar. Subjoined are 

 the results of the analyses of different specimens of blood in which Ber- 

 nard's process has been compared with my own. In every case where 

 my own process of analysis is put into practice I submit two samples of 

 the blood to examination, in order that the results may be checked, and 

 the same plan was adopted in some of the trials of Bernard's method. 

 In these instances, therefore, we have the results derived from the 

 analysis of four separate samples of the same blood, taken for examina- 

 tion at the same time : — 



Source of Blood. 



Sugar per 1000 parts. 



Gravimetric process. 



Bernard's volumetric 

 process. 



I. From bullock killed by Jewish 

 method 



II. From bullock killed by Jewish 

 method 



III. From bullock killed by Jewish 



method 



IV. From bullock killed by Jewish 



method 



V. From bullock killed by pole- 

 axe 



VI. From jugular vein of dog, 

 instantly after death 



VII. From sheep 



VIII. From case of severe diabetes, 

 obtained by cupping 



•588 



f a. -510 1 



[b. -489/ 



a. -515 



b. -535 



a. -698 



b. -709 



a. 1091 



b. 1-097 



795 

 811 



•509 

 •526 



f a. -795 1 

 \b. -811/ 



a. 4-990 



b. 4-951 



\ -588 

 j (mean 



•499 

 (mean 



•525 

 (mean 



•703 



(mean 



1-094 



(mean 



•803 

 (mean 



•517 



(mean 



4-970 



(mean 



•975 



a. -609"! -624 



b. -640 J (mean). 



1-025 



•869 



a. 1-311] 1-357 



b. 1-403 j (mean). 



•800 



•761 



a. 5-000 j 4-852 

 , b. 4-705 j (mean). 



The gravimetric process of analysis to which I have referred consists 



