SCIENCE. 



395 



In seven of the first seventeen cases the percentage of 

 reducing substance calculated by the above formula, ex- 

 ceeded that given by the copper test, and by a mean 

 amount of 0.539. 



In ten of them it fell short, and by a mean amount of 

 0.938. This method, therefore, can be relied upon to 

 give results which do not vary from the copper test ex- 

 cept by a small amount. 



Not much more in the way of accuracy can be claimed 

 for the copper test itself. 



In Nos. 18, 19 and 20 we have again the cases where 

 the high specific gravities vitiate the results of the calcu- 

 lation. 



CORRECTION FOR SPECIFIC GRAVITY. 



I next proceeded to find out a method for correcting 

 the reading of the polariscope for variations, caused by 

 changes in the speofic gravity of the specimens. First 

 I determined the percentage of water in glucose of differ- 

 ent specific gravities ; following are the results : 



I. 



Sp. gr. = 



Weight taken = 



Loss = 



Per cent H 2 = 



Sp. gr. 



Weight taken 

 Loss 



Per cent H 2 



Sp. gr. 



Weight taken 

 Loss 



Per cent H 2 



Sp. gr. 



Weight taken 

 Loss 



Per cent LUO 



Sp. gr. 



Weight taken 

 Loss 



Per cent HUO 



1.440 



5.51 5. in Pt. dish. 

 0.35, at 170 , 2 hours. 

 o.35 + 5-5I5 = 6 .37- 



II. 



I-43 1 



5.86 



o53. 

 o.53 



170 1 - 



2 hours, 

 j = 9-°5- 



Table V. 



III. 



1.409 

 4.038 



0.622, 170°, 

 15.40 



IV. 

 1.4 1 6 

 4.425 



0.525, 170° 

 H-93 

 V. 



i-4i7 



8.639 



1. 091, 170 

 12.70 



VI. 



3 hours. 



2 hours. 



hours. 



Sp. gr. 

 Weight taken 

 Loss 



Per cent H 2 



SOLID GRAPE SUGAR. 



I.463 



7-215. 



O.61 



9.29 



170 , 3 hours. 



These data are scarcely sufficient to establish a rule 

 for correction for variations in specific gravity, but it ap- 

 pears from them that the formulas will not vary much 

 from the following : 



The rule, 53 divisions = 53 per cent, seems applicable 

 to samples in which the percentage of H 2 is 12 to 14, 

 and of which the sp. gr. is from 1.409 to 1.414. For each 

 variation of 0.001 in the specific gravity, the percentage 

 of HaO varies about 0.3. 



Thus if we take the two extreme cases, viz. : 6.37 and 

 15.14 per cent of H u O, we find the corresponding speci- 

 fic gravities to be 1.440 and 1.409, a difference of 0.031. 



The difference in the percentage of water is 9.03. The 

 quotient of 0.0903 -4- 0.031 = 3 nearly. 



Let us apply these data to the correction of Nos. 18, 

 19 and 20 in table IV. I give below these numbers and 

 also their corrections. 



Number. 



Scale. 



52 ■ 63 

 5 6 -53 



53 7° 



% Reducing 

 Substance 

 by Cal- 

 culation. 



53 46 

 48-59 

 52.IO 



Same 

 Corrected. 



55-83 

 55 ■ 17 

 56.55 



Same by 

 Cu. Sof. 



56.81 

 54.60 

 51-14 



The above corrections were based on the supposition 

 that 53 divisions of the scale correspond to 53 per cent 

 reducing matter, when the sp. gr. — 1.409, and the per- 

 centage of water 15. 



We may therefore construct the following provisional 

 formula; for estimating the correction to be applied to the 

 reading of the scale when the sp. gr. of the specimen 

 varies much from 1.409. 



Let a = reading of scale. 

 " a' — corrected reading. 

 " e = sp. gr. of the sample. 

 Then a' = a — 3 a (e — 1. 409), when the sp. gr. is greater 

 than 1.409, and a' = a + 3 a (1.409 — e), when e is less 

 than 1.409. 



I next propose to undertake some investigations to 

 show the nature and number of the optically active prin- 

 ciples present in glucose. 



THE UNITY OF NATURE. 

 By the Duke of Argyll. 

 X. 



THE ORIGIN OF RELIGION CONSIDERED IN THE LIGHT 

 OF THE UNITY OF NATURE. 



( Concluded. ) 



In the beginning of this chapter I have observed how 

 little we think of the assumptions which are involved in 

 putting such questions as that respecting the origin of 

 Religion. And here we have come to a point in our in- 

 vestigations at which it is very needful to remember 

 again what some of these assumptions are. In order to 

 do so let us look back for a moment and see where we 

 stand. 



We have found the clearest evidence that there is a 

 special tendency in religious conceptions to run into de- 

 velopments of corruption and decay. We have seen the 

 best reasons to believe that the religion of savages, like 

 their other peculiarities, is the result of this kind of evo- 

 lution. We have found in the most ancient records of 

 the Aryan language proof that the indications of religious 

 thought are higher, simpler, and purer as we go back in 

 time, until at last, in the very oldest compositions of hu- 

 man speech which have come down to us, we find the 

 Divine Being spoken of in the sublime language which 

 forms the opening of the Lord's Pra^ er. The date in ab- 

 solute chronology of the oldest Vedic literature does not 

 seem to be known. Professor Max Miiller, however, con- 

 siders that it may possibly take us back 5000 years. 1 

 This is probably an extreme estimate, and Professor Mon- 

 ier Williams seems to refer the most ancient Vedic 

 hymns to a period not much more remote than 1500 

 B. C. s But whatever that date may be, or the corres- 

 ponding date of any other very ancient literature, such as 

 the Chinese, or that of the oldest Egyptian papyri, when 

 we go beyond these dates we enter upon a period when 

 we are absolutely without any historical evidence what- 

 ever, not only as to the history of Religion, but as to the 



1 Hibbert Lectures, p. 216. 

 a "Hinduism," p. 19. 



