Intelligence and Miscellaneous Articles. 79 



phosphate of sesquioxide of manganese, ascribed the purple colour 

 of the liquid obtained in Crum's method of testing for manganese, 

 by boiling any manganese compound with nitric acid and binoxide 

 of lead, to the formation of sesquinitrate of manganese, and not, as 

 Crum had done, to the formation of permanganic acid. 



The author has shown in the most decisive manner that Crum's 

 view is correct. 



When a solution, not too dilute, of permanganic acid or its potash 

 salt placed in a glass with parallel sides is brought into the solar 

 spectrum, it produces a very powerful absorption of greenish and 

 yellowish-green light. Solutions of sesquiphosphate of manganese 

 exhibit the same deportment. But if the latter is more and more 

 diluted, the obscurity in the middle of the spectrum gradually dis- 

 appears without showing any absorption hands ; while dilute solutions 

 of permanganic acid exhibit five distinct absorption bands, of which 

 the first feeble one (starting from red) lies more towards the Fraun- 

 hofer's line D, the second dark one is in the middle between C and 

 b, the third, equally obscure, lies upon E reaching to b, the fourth is 

 between b and F, and the fifth and feeblest is in F. These bands are 

 especially distinct when the spectrum is taken on a paper screen. 



Sesquichloride and sesquisulphate of manganese show quite a 

 similar deportment to the phosphate, excepting that there are new 

 absorptions in blue and violet. 



The liquid, prepared according to Crum's method, shows the five 

 absorption bands in the most distinct manner. — Zeitschrift fur 

 Chemie, No. 3, 1864. 



ON THE CONSTRUCTION OF DOUBLE-SCALE BAROMETERS. 



To the Editors of the Philosophical Magazine and Journal. 

 Gentlemen, 

 In a short paper which you did me the favour to insert in your De- 

 cember Number, it was remarked that there was reason to fear that 

 double-scale barometers were not always properly graduated in this 

 country. I was then unaware of the extent to which barometers 

 with incorrect metrical scales have recently been manufactured by the 

 most eminent firms in London. I have in my own possession a 

 standard barometer, two mountain barometers, and a mountain 

 aneroid by four of our first makers, and in each of these the metrical 

 scale is erroneous, while a mountain aneroid by Secretan of Paris is 

 affected by a similar error in the English scale. My friend Mr. 

 Tuckett of Bristol, who has been investigating this subject simulta- 

 neously with myself, informs me that all his instruments are like- 

 wise inaccurate. 



The nature of the error may be explained as follows : — 

 Suppose a standard brass yard, divided to inches, at the normal 

 temperature of 62° F., to be laid beside a standard brass metre di- 

 vided to millimetres at the normal temperature of 32° F., with their 

 zero-points coincident. Then by Guyot's Tables for converting 

 inches into millimetres, at their respective normal temperatures, each 

 inch of the English standard will correspond with 25'39954 millims. 

 and 30 inches with 761*986 millims., or 762 millims. very nearly. 



