4:02 T. A. Jaggar, Jr. — Microsclerometer, for 



ec hanical structure, the second atomic structure. Again, if 

 a mass of amorphous carbon of low specific gravity is heated 

 in a crucible, crystalline graphite needles are formed, of some- 

 what higher specific gravity and hardness ; if enclosed in a 

 globule of molten iron, which expands suddenly on solidifica- 

 tion, and so subjects its inclusions to great atomic pressures, 

 Moissan* has shown that the carbon crystallizes in the form of 

 carbonado and diamond, of very high specific gravity and 

 hardness. The substance iu this case is a single element 

 throughout, its atom has retained a constant weight but the 

 molecule has continuously increased in weight under pressure, 

 hence the number of atoms in the molecule has increased with 

 the hardness. This accords with the determinations by Bot- 

 tonef that the hardness of metals varies as the specific gravity 

 divided by the atomic weight, which quotient, as Turner^ has 

 pointed out, varies directly as the number of atoms in a unit 

 space. 



Atomic structure in the element ; atomic and molecular 

 structure in the crystalline compound ; atomic, molecular and 

 sub-microscopic structure in the amorphous or crypto-crystal- 

 line substances; atomic, rr^ecular (sub-microscopic) and 

 mechanical sti the • 'ystalline or fragmental aggre- 



rdness as measured by physical 

 rative results of scientific value 

 portant that definite crystals or wholly amorphous sub- 

 stances be selected. If this is done, abrasion tests, delicately 

 conducted, may give very valuable data concerning the inti- 

 mate structure of solids : Exnerg has shown very interesting 

 relations between the directions of resistance to abrasion and 

 crystalline form : Pfaff|j has barely touched upon a wide field 

 of research in his experiments on the mean hardness variations 

 of minerals in isomorphous series, and the relation of the cohe- 

 sion constants to the other physical properties. And, finally, 

 as an aid to the differentiation and determination of the crys- 

 talline minerals, there is no reason why a thoroughly defined 

 method may not give very constant results. 



Former instruments have had three chief sources of error : 

 (1) personal variability due to using " visibility" as determi- 

 nant; (2) inequalities of mineral surface ; (3) undefined details 

 of instrument. To eliminate (1), the depth of abrasion should 

 be definite and measurable : to eliminate (2), the surface should 

 be artificial and defined, and the boring method, where only a 

 very small portion of the surface is initially touched, should 

 be used: this, at the same time, gives a mean value for all 

 directions in the surface ; to eliminate (3), every part of the 



*Comptes Rendus, 1894 and 1895. 



f Loc. cit. % Loc. cit. § Loc. cit. || Sitz. k. k. Bayer. Akad., 1884, p. 255. 



