'62 THOMAS STEERY HUNT ON A NATURAL SYSTEM IN 



the various orthophosphates aud ortharseniates of sodium with 12HoO, have, according to 

 Playfair aud Joule, equivalent volumes of from 233 to 235, while ferrocyauid of potassium 

 gives 230, lactose 234, aud piperiue (with a density of 1.244) 4*76, or about double these 

 numbers. Other species, as it was poiiated oiit, have apparently an equivalent volume of 

 430, and still others about 200, or some multiple of this number. Whether the weights 

 thus assigned to various silicates and carbon-spars might represent their chemical eqriiva- 

 lents or some portion thereof, they, in any case, served to show the relative condensation 

 of matter in the different species compared. 



§ 19. This subject was continued a few months later in a paper read at Washington 

 in May, 1854, before the American Association for the Advancement of Science, entitled 

 "Illustrations of Chemical Homology." ' Therein were reviewed and reaihrmed the 

 teachings of the two papers of 1853, while the principles of homology were farther 

 exemplified, and it was maintained that homologies may exist alike between species differ- 

 ing by «(M.Oo) and nÇR.D.,), and even between those related species which differ in 

 the proportion of silica, so that the ratio between silica and bases has but a specific A^akie. 

 It was further contended that the water contained in a great many hydrated species often 

 described as altered silicates, was to be regarded as not of subsec^uent introduction but an 

 original and essential element of the species, as is admitted to be the case in the zeolites. 



§ 20. In the second i^aper for 1853 was considered the question of chemical notation 

 and formulas, which was farther illustrated in the paper of 1854. At this time the question 

 of the atomicities of the elements had not yet been discussed, and the distinction between 

 univalent and bivalent metals, suggested by Cannizaro in 1858, was unrecognized. The 

 symbols then used for both of these stood for one atom, or for the proportion which 

 in the so-called protoxyds is united with eight parts by weight of oxygen. In sesquioxyds, 

 like alumina, however, recognizing the trivalent character of ALO, {2T-I-24), it was by 

 the writer regarded as corresponding to three atoms of oxyd of alumiuicum = 3alO. 

 Silica, which, following Berzelius, was then generally written SiO.j (21+24) became 3siO. 

 With this notation were constructed atomic formulas, the elements now regarded as 

 diatomic being confoirnded with monatomic elements, and, like these, represented by 

 capital letters. Thus the common atomic formula for bisilicates, as given above, was 

 written n{siMO.^) and spodumene, n = 30, was made si^„0,,j(al,,,LiiNa.)0.,„. Similar atomic 

 formulas are still employed in these pages, using, however, small letters to represent 

 an atom of any element, whether univalent, like sodium or chlorine ; bivalent, like 

 calcium or oxj^gen ; trivalent, like boron ; c[uadrivalent, like silicon, titanium and carbon ; 

 or sexvalent, like the double molecule of aluminum. The above general formula is thus 

 now written w(si.m,03), and that given for spodumene (si^uaL,^lijna.)Oi,„. In order to dis- 

 tinguish the atorn of ferrosum = jFe, from that of ferricum = jFe, the former is written 

 fe, and the latter fi, while manganicum, corresponding to manganic sesquioxyd, is mni. 



§ 21. The M in the general formtila M.O., employed in 1853, was thus made to repre- 

 sent an atom either of protoxyd or sesquioxyd. and in 1854 a farther generalization was 

 attempted. The boric, titanic, tantalic and niobic anhydrids were reduced to the same 

 atomic formula as silica, and moreover, in view of the variations in the silica-ratio in 



' Proc. Amer. Assoc. Adv. Science, 18.54, pp. 237—247 ; also in abstract Amer. Jour. Science for Sept, of the 

 .same year, and noticed, with extracts, in the author's Chem. aud Geol. Essays, p. 438, et srq. 



