tHEORIES OF CRYSTALLOGRAPHY. 83 



top is not fymmetrical with the bottom. This latter fpecies of taws of the 

 fymme(ry appears, by general obfervation, to be exclutively cryftals, and 

 appropriated to cryftais that become ele<5lrical by heat; that theory thence 

 is to (ay, which being expofed to the heat of the fire, or '"'^^"^""S* 

 plunged into hot water, acquire th j eledric power. Thefe 

 cryftals, the tourmaline, for example, acquire a pofitive elec- 

 tricity on one fide, while on the fide diametrically oppofite 

 their electricity becomes negative ; and all obfervations hi- 

 therto made give us reafon to conclude that thefe fides are 

 never fymmetrical, and are always produced by different or 

 fewer laws of decrement. " Hence,^' fays the Abb6, *• by 

 mere infpedlion it is eafy to point out which is the fide that 

 will give the pofitive and which the negative eledlriclly." 

 (Vol. i. p. 237.) 



The aftonifiiing variety in the cryftalline forms leads us 

 naturally to afl<:. What can be the caufe of this variety ? This 

 queftion has not been treated by the Abb^ : allow me. Sir/ 

 to fubmit a few ideas on the fubje6l for the opinion of mathe- 

 maticians. 



Firft caufes, I repeat, are not the obje6t of this difcuflion. 

 I ftate the queftion thus : Why does the fame fubje<5t cryftal- 

 lize in fuch a variety of forms, always fymmetrical and always 

 terminated by planes? 



The folution of this queftion feems to require three condi- 

 tions: 1ft, That the particles of the fubftance diflx)lved in the 

 fluid all leave the ftate of reft at the fame inftant, to form the 

 cryftal by their aggregation : 2dly, That, while thefe particles 

 are in the ad of drawing near to each other, no foreign power 

 (hall imprint on them any other motion than a common motion, 

 whether it be in a ftraight line, or rotary round their common 

 centre of gravity : 3dly, That the particles all arrive at the 

 ftate of reft at the fame inftant, which takes place when the 

 a6t of cryftallization is finithed. The fecond condition is ne- 

 ceftary, and infers the firft and third. The natural confe- 

 quence of thefe conditions will be, that the aggregation of the 

 particles will only take place conformably to a law ading 

 equally on all of them, whatever may be the law. 



Since they all leave the ftate of reft at the fame inftant, they 

 are in equihbrio previous to ihat inftant. Since they all arrive 

 at the ftate of reft at the fame inftant, they are in equilibrio 

 after that inftant : but when particles that are aded upon by 



G 2 no 



