Intelligence and Miscellaneous Articles. 531 



In order to prove this I constructed one specially for deter- 

 mining the densities of the worts of Belgian beers, which vary 

 only between 1*00 and 1*06. The scale, extending over a length of 

 about 24 centimetres, is graduated 1-00, 1-005, 1-01, 1-015, . . . 1-06. 

 Between each two of these divisions are four large and five inter- 

 mediate ones, making nine equidistant divisions of about 2 millims. 

 The third decimal can thus easily be read with accuracy ; and the 

 fourth can be obtained with approximation amply sufficient for 

 practical purposes, as is proved by determinations made by the aid 

 of the apparatus and rigorously controlled by the method of the 

 specific-gravity flask. 



This densimeter has many advantages which we may point out. 



I. All ballast is suppressed, whether shot or mercury, aud thus 

 the apparatus is less heavy and less subject to fracture. 



II. Correction for capillarity is dispensed with, seeing that we 

 always work in distilled water, which can always be taken in the 

 same conditions. 



III. The volume of the liquid with which we work need not be 

 measured, being constant. (In this respect the instrument has 

 the advantage over other densimeters which depend on the same 

 principle, such as those of Eousseau, Paquet, Laska.) 



IV. The density of even a small quantity of liquid may be 

 determined. 



V. It is convenient for determining viscous liquids. 



VI. The temperatures of the liquid and of the distilled water do 

 not differ during the operation. 



VII. Calculation shows that the results obtained are indepen- 

 dent of the temperature worked at, at least for considerable 

 differences. Allowing for the expansion of the liquid, that of the 

 apparatus, and that of the water of immersion, it will be seen that 

 these effects almost exactly compensate one another, especially if 

 the liquid and the water are taken at the same temperature. Expe- 

 riment shows that even with the instrument mentioned above, 

 which has been graduated at 15° C, we may take the density 

 indifferently from 8° to 9° C. to 19° or 20° C," without the results 

 differing by more than 1 to 2 units in the fourth decimal place. 



VIII. The instrument can, lastly, be easily cleaned. 



The apparatus is also constructed for liquids less dense than 

 water. The division 1-00 is then at the top of the scale; and 

 putting the division 1-00 at the middle of the scale, we can wdth 

 one and the same instrument, within certain limits, determine 

 densities above and below that of water, especially if we are 

 satisfied with a result correct to the second decimal and approxi- 

 mately so to the third. — From a separate impression from the 

 Arch. Sci. Phys. et Nat. t. xxxiv. p. 409, communicated by the 

 Author. 



