March 17, 1893. J 



SCIENCE. 



Hi 



2. If a standard litre of distilled water, free from lair, be 

 weighed in London against brass weights in air at 62° F., barom. 

 30 in., the result will be 1539-'. 6 grains (p. 103); and the weight 

 of 1 cubic inch (of 1890) of water under the same conditions is 

 252.2S6 grains. Hence the volume of the -'London" litre (of 



1890) is ^539«-6 



253.286 



3. The volume of the 



60.9947 cubic inches. 



15398.6 



61.036384 cubic inches. 



'London ' litre (of 1834) is 



15398.6 



adopted by Mr. Clark is 61.0364. 



4. The value of 



253.286 



5. The weight of 1 kilogram of distilled water, free from air, in 



vacuo, at 62° F., is 



15432.35 



15415.0851 grains, and the loss by 



1.00113 



weighing in air is 16.491 grains. The weight of the litre in Lon- 

 don at 63° F.. barom. 30 in., is thus 15398.594 grains; and this, 

 divided by 353.386, gives 61.03636 cubic inches as the volume 

 of the London litre (of 1890). 



6. On the basis of 61.03636 cubic inches per litre, the number 

 1728 



of litres in 1 cubic foot is 



= 38.31104. 



61.03636 



7. On the basis of 61.0364 cubic inches per litre, the number 

 of litres in 1 cubic foot is 38.310975. 



1738 



8. The value of 

 28.3110. 



Cu. in. per litre 



adopted by Mr. Clark is 



B.— CDBE DECIMETRES. 



1. The English metre is 39.37079 inches. Hence the English 

 cube decimetre is 61.027051 cubic inches. 



2. The weight of 1 cubic decimetre of distilled water, free 

 from air. weighed in air against brass weights, at 4*^ C, bar. 30 

 in., is 15413.5 grains; and the weight of 1 cubic inch under simi- 

 lar conditions is 352.568 grains. Hence the volume of the stand- 



1 KA-l 'A f; 



ard cubic decimetre (English 1890) is i^!^i£:- = 61.037139 cubic 



353.568 

 inches. 



3. The value of (3.937079)=, adopted by Mr. Clark, is 61,0270 

 cubic inches. 



4. The U. S. metre is 39.37 inches. Hence the U. S. cube deci- 

 metre is 61.033377953 cubic inches. 



5. On the basis of 61.037051 cu. inches per cube decimetre, the 



1728 



number of cube decimetres in 1 cubic foot is 

 28.31531. 



6. The value of 



1738 



61.037051 

 adopted by Mr. Clark, is 



cu. in. per cu. dec. 

 38.3153. 

 7. The number of U. S. cube decimetres per cubic foot is 

 1738 



61.023377953 



-= 28.31702. 



C— CONVEESION VALUES. 



1. " Cube centimetres into cube decimetres (litres)" — divide 

 by 1000 (p. 17). 



2. Cube centimetres -'into litres" — divide by 1000.05 (p. 17). 



3. Cube decimetres "into litres" — multiply by 1 (p. 24). 



4. " Kilogram = 1000 grammes = 1 litre, or 1 cube decimetre 

 water, 4° C. Miller, in 1856, found the kilogram - 15482.319 

 grains in vacuo. It was originally intended to be the weight of 

 a cubic decimetre of water at maximum density in vacuo. It is 

 now a definite mass of plat num and is slightly heavier than the 

 cubic decimetre of water" (p. 50). 



5. "Cube metres or steres (=1000 litres very nearly) into 

 litres" — multiply by 1000 (p. 61). 



" Cube metres or steres into cube decimetres" — multiply by 

 1000 (p. 61). 



"Cube metres or steres into cube feet" — multiply by 

 35.31658 (p. 61). 



"Cube metres or steres into cube inches" — multiply by 

 61037.05 (p. 61). 



6. "Kilograms (or litres) of water into cube inches" — mul- 

 tiply by 61.170 (p. 92). 



■ multiply by 28.311 

 -multiply by 



cu. inches. 



"Cube feet of water into litres, 63'^ F." 

 (p. 93). 



" Cube feet of water into kilograms 62° F."- 

 38.249 (p. 93). 



From this summary it will be seen that Mr. Clark's book and 

 letter present us with quite an extended range of choice for the 

 value of a litre, viz. : — 



Standard litre (1890) 61.036273 



decimetre 61.037051 



(weighed in air) 61.037139 

 London litre (1890; 61.03636 



" (1824) 60.9947 " 



Clark " (1890) 61.036384 



" 61.0364 



" decimetre 61.0370 " 



6103705 

 U.S. " 61.033377953 



"kilogram" (in vacuo, 4° C, 1890) 61.104666 " 



(in air 63° F.) 61.170 



and, in addition to these, I may quote the following from Thl.le 1 

 of the before-mentioned article on " Fuel-Gas Values," viz. : — 

 Authority. Cu. ins. in litre. 



U. S. Dispensatory. 16th ed. 

 G. Gore, LL.D , F.R.S. 

 Professor V. B. Lewes, F.C.S. 

 Professor J. D Everett, F.R.S. 

 Trautwine (said to be U. S. Standard) 



Haswell (said to be by Act of Congress) 



61.0280 



61.024 



61.034 



61.023 



61.0254 



61.034435 



61.032 



61.03534 



61.0367 



61.03709 



61.02433 



64.99008 



Gmelin 



W. Crookes, F.R S. 



Thomson and Tait 



S. A Ford 



The suggestion made by Mr. Clark that these discrepancies 

 may for the most part be explained by the ditference between the 

 1834 and 1890 standards is obviously insufficient if the difference 

 he refers to be that of the cube inch value ; for as the 1834 value is 

 60.9947 it clearly was not adopted by the authorities above quoted. 

 Some other explanation is, therefore, required; and as so con- 

 summate an authority as Mr. Clark appears unable to advance 

 one, I may perhaps be allowed to hint that the cause of the vary- 

 ing values is to be found in sheer laxity of calculation. 1 know 

 that so commonplace a theory is rather shocking, and I duly 

 blush as I advance it; but, really, when I find Mr. Clark himself 

 deliberately adopting the value 61,0364 as the quotient of 



^^^^^•^ and adopting it as the basis of his book, whereas the true 



352.2fe6 



quotient is 61.0B62!:4, or, if four places of decimals be used, 



61.0363, I may plead for pardon with some assurance of the same 



being accorded. The example here cited is even still more to 



the point; for the value 15398.6 is adopted by Mr. Clark as the 



result of the calculation l^lSSj^ _ 16.491, whereas the true result 



is 15398.594 and this divided by 253.286 gives 61,03636. 



But let it not be imagined that I make these remarks in any 

 fault-finding or critical spirit. I am too conscious of my own 

 short-comings to be willing to sit in the seat of judgment. In the 

 before-mentioned table, for example, I derived Mr. Clark's second 

 value of "cubic feet in 100 litres " from his figure of 61,04 cubic 

 inches per litre. The calculation was, of course, ^^—=3 5334; 



and yet, when I corrected the proof of the article, I inadvertently 

 allowed the value to appear as 3.5323. So I must ask my scien- 

 tific brethren to understand that my observations are not intended 

 as any disparagement of the " Dictionary of Metric Measures'" or 

 as casting any adverse reflection upon the other text-books I have 

 quoted, all of which I regard as admirable examples of scientific 

 work and as trustworthy as reasonable mortals can expect them 

 to be. 



And so we come back once more to our question, Where, 

 after all, is the litre ? Our puzzle-picture turns out to be of a 

 kaleidoscopic variety and appears in a different aspect to every 



