Tourmaline and other Minerals 'when exposed to Heat, 1 39 



their maximum intensities measured. Two series of experi- 

 ments in a direct order, and two reversed, all gave the same 

 order of intensity for these specimens, which, instead of bear- 

 ing any direct proportion to the areas, as might have been ex- 

 pected, where the lengths were equal, gave the following ar- 

 rangement* in the order of intensities : 1, 2, 5, 4, 3, the 

 areas following the natural order of the numbers. Other series 

 being taken with sets of crystals 1*2 and 1*8 inches long, gave 

 similar indications of irregularity f; but the area of section has 

 so far a general influence, that where the differences are con- 

 siderable, the thickest crystal has almost universally the great- 

 est power. The relative forces are so connected that we can 

 hardly impute the irregularities to any general law : the dif- 

 ferences, as I shall immediately illustrate by reference to an- 

 other class of experiments, must in all probability be attributed 

 to a variable structure in specimens of the same mineral, as 

 well as in those of different species. 



I took a crystal l£ inch long, and carefully determined 

 the intensity of its electricity, which, by a mean of three expe- 

 riments, gave 45° of deviation. I immediately broke it at one 

 fourth of its length from one end : the two portions being then 

 heated, and their intensities determined each three times, the 

 mean of the larger portion gave a deviation of 47°, of the 

 smaller 43°, the mean of which gives precisely the original 

 force. As far as intensity goes, the diminution of length would 

 not therefore appear to be favourable to the development of 

 electricity. With a view of procuring through a larger range 

 of dimension the influence of length alone, I selected a series 

 of tourmalines whose sections were as nearly equal as possible, 

 the diameter being about T Vth inch, and one of which was the 

 very long crystal before mentioned. This experiment was 

 made with great care ; a direct and reversed series were taken, 

 and several of the determinations independently repeated. 



* The best pair of series gave, 



No. 1, Deviation 115° 



•5 



3, 



4, 26 



5, 39-5 



Nos. Intensity. Intensity. 



1*2 long. 1*8 long. 



1 (thickest) 82° 54° 



2 775 40 



3 50 34 



4 575 355 



5 65 



6 (thinnest) 34 



T2 



