Thoughts on Magnetifnt. 9t 



3. By cryftalization I uncferftand that power by which the integrant particles of any folid 

 poffeffing fufiicient liberty of motion unite to each other,' not indifcriminately and confufedly, 

 but according to a peculiar uniform arrangement, fo as to exhibit in Its laft and moft perfe£t 

 ftage regular and determinate forms. 



4. This po%ver is now known to be poflefled by all folid mineral fubflances. 



5. The forms which cryftals, even of homogeneous fubftances, exhibit, are often very 

 numerous ; however in moft cafes they may be reduced to a few primordial forms, which, 

 as Abbe Hauy has lately experimentally proved, are derived from certain original forms 

 appertaining to the minuteft particles of their concretion. 



6. The aflemblage of thefe ultimate particles into vlfible aggregates, fimilarly arranged, 

 neceflarily requires that one of their furfaces ifhould be attra£live of that particular furface of 

 the other, which prefents a correfponding angle, and repulfive of that which prefents a dif- 

 ferent angle, otherwife the various regular rhomboidal and other polygon prifms and pyra- 

 mids, which cryftals prefent us, could never exift; confequently the minuteft prifm, being 

 once formed, could never be prolonged if one end of fuch prifms were not attractive, and 

 the other repulfive of the fame given furface. 



7. Hence it has been obferved that cryftalization never takes place in the middle of any 

 folution, but always begins at the furface or on the bottom or fides of t|)e vefTels that contain 

 it, for the particles in the middle of the folution being confufedly mixed with each other, 

 and exerting their repulfive as freely as their attra£live powers, the one conftantly counter- 

 afting the other, no fenfible accretion of a regular kind could take place, whereas the 

 repulfive power of the uppermoft particles, or of thofe that reft on the fides or bottom of the 

 veflel, is reftrained and impeded. 



8. The repulfive power of cryftalizing fubftances alfo appears in many other inftances 

 (of the attradive no doubt has ever been formed.) Thus if faturate folutions of nitre, 

 common fait, and tartar vitriolate be mixed and fet to cryftalize, each will cryftalize a part, 

 which could not happen if the particles of each of thefe falts did not only attradl: their 

 fimilar homogeneous, but alfo repel thofe of a dilFerent fpecies, otherwife the mere cafual 

 circumftance of greater proximity to one than to the other would impel them to unite indif- 

 criminately. Again, if a faturate folution of allu;n be mixed with a turbid mixture of clay, 

 and abandoned to infenfible evaporation, after fome time the clay will fubfide and form a 

 dry mafs, but in the interior of this mafs large regular cryftals of allum will be found, 

 whofe component particles muft, to reunite, have difplaced and repelled the particles of clay 

 with which they were furrounded. 



9. If to a faturate folution of a fait that difficulty cryftalizes, a cryftal of a fait of the fame 

 fpecies be inferted, the whole folution will foon be brought to cryftalize, as the cryftal 

 inferted attrafts the particles diflblved, by its different furfaces ; But if a fait of a different 

 nature be inferted this will not happen, cryftalization will not be promoted. 



10. If to a folution of 2 parts nitre, and 3 parts Glauber's fait in 5 parts water, a cryftal 

 of nitre be inferted, the nitre alone will cryilalize; or if inftead of nkre a cryftal of Glauber 



N2 be 



