ICi ANNUAL OF SCIENTIFIC DISCOVERT. 



that the atomic weight of copper needs correction, and should be 

 doubled to 63.4, in which case the quotient would be 7.204:, and 

 it would then fall amono^ the other non-masrnetic metals, 



od. The quotients are the smallest for those metals whicli are 

 the most permanently magnetic^ even at high temperature, and 

 vice versa. 



4th. As cooling increases by contraction the number expressing 

 the specific gravity, it will consequently decrease the quotient 

 obtained by using this increased specific gravity as a divisor, in 

 perfect accordance with the fact that cooling increases the para- 

 magnetic property. 



5fh. As, inversel}^, heating decreases by expansion the specific 

 gravit}^ it will increase this quotient, in accordance with the fact 

 that heat diminishes paramagnetism, and finally destroys it in all 

 metals, with the single exception of cobalt, which has the smallest 

 quotient of all, and consequently can stand some increase. 



6th. The experiments of Faraday on diamagnetism and para- 

 magnetism, with very powerful electro-magnets, have proved that 

 palladium and platinum are the strongest paramagnetic next to 

 the first five in the above list; they have in my list, also, the 

 smallest quotients connected with them. 



7th. In the same way as diamagnetism is the opposite of para- 

 magnetism, the larger quotients in the above table belong to dia- 

 magnetic bodies, as, for instance, mercury, antimony, and bismuth. 

 The last is the strongest diamagnetic substance experimented 

 upon, and possesses the greatest quotient in the above table. 



8th. If we were able to cool the other metals so as to increase 

 their specific gravities to such a degree as to have a decided effect 

 on the amount of this quotient, we might perhaps succeed in dis- 

 covering in several of them paramagnetic qualities by means of 

 Faraday's apparatus. 



9th. Heating decreases the paramagnetic qualities, with the spe- 

 cific weight, and consequently increases the quotient. That it 

 may do this to such a degree as to make the body diamagnetic, is 

 proved in the case of oxygen gas, which, when cool, is paramag- 

 netic like iron, and, when hot, diamagnetic like bismuth. 



10th. That this relative distance of the atoms (upon which, of 

 course, the specific gravity of bodies depends) is closely related to 

 their magnetism, is again proved by their crystals, which are al- 

 waj's less dense in the direction of their optical axis, and expand 

 by heat more in one direction than in another ; and by Pluecker, who 

 has demonstrated that they are diamagnetic in the direction of 

 their optical axis, or of the longest axis of crystallization. In some 

 of these crystals this action is so strong that they are influenced 

 by the magnetism of the earth ; as, for instance, a properly cut 

 crystal of kyanite (a dense silicate of alumina), when suspended 

 on an axis, will behave like a compass needle, and may be used 

 as such, — a fact little known, but worth knowing. — Amer. Mining 

 Journal. 



