106 



CHEMISTRY. 



as a non-nitrogenous, valueless residue and the 

 indication of what may be their function in the 

 physiology of the plant. 



Prof. E. W. Morley described at the American 

 Association his investigations of the volumetric 

 composition of water, together with the appa- 

 ratus employed and the results. The apparatus, 

 with 300 feet of glass tubing, occupying parts of 

 two rooms, includes provision for producing hy- 

 drogen, arrangements for purifying it, a globe 

 in which part is collected to be weighed, another 

 where the gas is stored for analysis, and three 

 other parts in which the analysis is made. In a 

 second apparatus, identical with the first, the 

 volumes in which hydrogen and oxygen are made 

 to combine are measured, and the remaining ex- 

 cess is determined. From this was computed 

 the ratio of combination in 20 determinations. 

 The minimum value was found to be 2-00005, 

 the maximum 2-00047, and the mean 2-00023, 

 with a probable error of determination of one 

 part in 30,000. In a second paper, on the ratio 

 of the density of oxygen and hydrogen, Prof. 

 Morley said he had made two determinations of 

 the specific gravity of hydrogen, and reached the 

 same figure for the ratio as that given by Lord 

 Rayleigh, in England, namely, 15-884. This 

 value, combined with the preceding, gave 15-882 

 as the atomic weight of oxygen. 



Prof. Noyes narrated the results of four series 

 of six determinations each of the atomic weight 

 of oxygen, with apparatus devised by himself. 

 He had found the value to be 15-896. 



The seed of Calycanthus glaucus is found, by 

 the analysis of H. W. Wiley, to be rich in oily 

 albumenoids and sugar (of which it contains 

 twice as much as wheat), and poor in starch and 

 undetermined substances. The oil has a beauti- 

 ful faint-yellow color and a peculiar odor. The 

 seed also contains a poisonous alkaloid, which 

 Dr. R. G. Eccles, of Brooklyn, has named caly- 

 canthine. Attention was called to the poisonous 

 property of the seed by a letter from Mr. J. H. 

 H. Boyd, of Cagle, Tenn., relating how cattle 

 and sheep had been fatally poisoned by eating 

 the fruit of the shrub. The symptoms of the 

 poisoning resemble those of drunkenness, with 

 extreme nervousness. Sometimes the fatal issue 

 is immediate, sometimes it is delayed for three 

 or four weeks. The alkaloid is not very virulent, 

 is slightly soluble in water and very soluble in 

 ether and chloroform, while its salts are insolu- 

 ble in chloroform but very soluble in water. It 

 crystallizes from ether in feathery masses, which 

 form very rapidly. 



The results of systematic examinations by G. 

 J. Fowler and J. Grant of the influence of the 

 chief metallic oxides and certain unstable salts 

 on the decomposition of potassium chlorate by 

 heat may be summarized as follows: 1. Acid 

 oxides, such as V a 8 , WO S , and V 8 8 , cause the 

 evolution of oxygen at a much reduced tempera- 

 ture, with the formation of a metavanadate, 

 tungstate, or uranate. Chlorine is evolved in 

 large quantity in these cases, but the whole of 

 the oxygen of the chlorate is not liberated. 2. 

 Alumina acts in a similar way, but less ener- 

 getically. 3. Chromium cesquioxide causes the 

 evolution of oxygen at a lower temperature, with 

 the liberation also of chlorine. 4. The sesqui- 

 oxides of iron, cobalt and nickel, cupric oxide, 



and manganese dioxide cause the evolution of 

 oxygen at a comparatively low temperature, ac- 

 companied by only a small percentage of chlo- 

 rine, while the oxide is left but little altered at 

 the 'end of the experiment. 5. The monoxides 

 of barium, calcium, and lead cause no evolution 

 of oxygen when heated with potassium chlorate, 

 but the latter breaks up below its normal tem- 

 perature with the formation of potassium chlo- 

 ride and a peroxide. 6. In the presence of such 

 oxides as silver oxide and the peroxides of ba- 

 rium and lead, potassium chlorate acts as a re- 

 ducing agent. No oxygen is eliminated, but a 

 perchlorate is formed. 7. Oxides, such as those 

 of zinc and magnesium, are inactive. The au- 

 thors find that the physical condition of the 

 oxide is of importance thus copper prepared in 

 the dry way is almost inactive ; and further, that 

 certain substances, such as powdered glass, sand, 

 and kaolin, assist the decomposition, although 

 they apparently undergo no chemical change. 



Practically three methods are now available 

 for the determination of lithia in mineral wa- 

 ters ; the phosphate method (Mayer's modifi- 

 cation) ; the amylalcohol method of Gooch ; and 

 the fluoride method of Carnot. Bammelsberg's 

 method is somewhat similar to that of Gooch, 

 but has not proved very satisfactory in its ap- 

 plication. For all of these processes it is neces- 

 sary to obtain from some known quantity of the 

 water the alkalies as chlorides free from admixt- 

 ure with other bases, and in most cases a con- 

 siderable proportion of the sodium and potassium 

 salts, which usually predominate over those of 

 lithium, must be removed. In order to test 

 these methods upon water containing lithia, 

 samples of the best known and widely advertised 

 waters were purchased by E. Waller and sub- 

 mitted to examination. The results were some- 

 what surprising, and indicate that either the 

 original analysis, on the strength of which the 

 waters are sold, was erroneous, or, what is more 

 probable, that the proportions of lithium in thosfe 

 waters are liable to great fluctuations. The re- 

 sults were chiefly obtained by Carnot's fluoride 

 method, but were in several cases confirmed by 

 the use of other methods. In the Farmville 

 lithia water no lithium could be detected by the 

 spectroscope in moderate amounts, and only 

 traces in larger amounts. The reaction for lith- 

 ium in Buffalo water in considerable quantities 

 was more distinct. In the Londonderry water 

 the lithia reaction could be obtained without 

 great difficulty. Of all the waters examined, 

 purporting to be natural, the Saratoga Hathorn 

 proved to be the strongest in lithia. The lithia 

 waters manufactured and sold by Carl H. Schultz 

 were found to contain a little more lithia than 

 was claimed for them. 



The following tests are given by M. M. Grand- 

 val and Valser for the detection of linoleic acid 

 in oleic acid. The falsified acid is of a yellowish- 

 brown tint, paler than that of oleic acid ; its 

 specific gravity is higher. The liquid is more 

 consistent, and is not homogeneous, but gritty. 

 If falsified oleic acid is heated to 50 C., it takes, 

 when cold, a firmer condition, which becomes 

 more decided each time the operation is re- 

 peated. A mirror-like precipitate is produced 

 on shaking with alcohol, while oleic acid dis- 

 solves. If mineral oil, resin, or paraffin is mixed 



