740 



SCIENCE. 



[N. S. Vol. VIII. No. 204. 



NOTES ON INORGANIC CHEMISTRY. 

 The study of the influence of chemical 

 composition on the coefficient of expansion 

 of glass is one that has attracted consider- 

 able attention from both theoretical and 

 practical standpoints. According to Nature 

 an interesting resume is given by M. A. 

 Granger in the Moniteur Scientifique. In a 

 few cases only the expansion follows an ad- 

 ditive law proposed by Schott. A number 

 of substances, such as the oxids of lead, cal- 

 cium, manganese, aluminum and boron, 

 lower the dilation when added in small 

 quantities, but raise it when the proportion 

 is increased. Potash, soda, lithia, fluorspar, 

 lime or calcium phosphate raise the coeffi- 

 cient of expansion, but except in the case of 

 the last not more than 8 per cent, can be 

 added, as the glass either refuses to take up 

 more or else becomes devitrified and 

 opaque. Calcium borate, oxid of iron, 

 alumina and silica lower the coefficient of 

 expansion, alumina being especially active 

 in this respect. 



A SECOND series of experiments on the 

 action of water on metals is contributed to 

 the last Chemical Neivs by Robei-t Meldrum. 

 The action on iron was noticed last week 

 in this column. He finds that all waters 

 tested have action on copper. Seven feet of 

 one- sixteenth inch wire was used in each 

 case. In five months distilled water had 

 dissolved 0.055 parts per 100,000. Am- 

 moQia and carbon dioxid free water in 115 

 hours' exposure contained 0.1925 parts. A 

 lake water containing 0.0056 free ammonia, 

 0.0126 albuminoid ammonia and 1.756 

 chlorin dissolved in 24 hours 0.099 parts. 

 A water with no free ammonia, 0.001 al- 

 buminoid ammonia and 1.22 chlorin in 24 

 hours dissolved 0.023 parts copper. A town 

 water supply with 2.07 chlorin and 3.0 or- 

 ganic matter dissolved in 94 hours 0.0825 

 copper, all in parts per 100,000. Sludge 

 from a water tube boiler in use for some 



years contained 0.006 per cent, copper, 

 showing the solvent action on copper and 

 brass fittings. ISTo zinc is mentioned as 

 being present, though it is known that some 

 waters exercise a decidedly solvent action 

 upon the zinc in the brass, affecting but 

 slightly the copper. 



Experiments on lead were cari'ied out by 

 exposing the water in pieces of new lead 

 pipe closed at one end. Two waters were 

 tested : one (A) of a permanent hardness of 

 3.2° and a total hardness of 3.3° ; the other 

 (B) of permanent hardness 5° and total 

 hardness 18.6°. In four hours A had dis- 

 solved 3.97 parts per 100,000 and B 0.049. 

 When containing a small amount of carbon 

 dioxid the solvent action was unchanged, 

 but when almost saturated with carbon 

 dioxid the solvent action of A was after the 

 first half hour greatly increased. When 

 saturated with calcium bicarbonate the sol- 

 vent action was greatly decreased and 

 when water A was agitated with calcium 

 carbonate and then filtered, it ceased to 

 have any solvent action. These experi- 

 ments bear out the generally accepted view 

 that hard waters take up little lead from 

 lead pipes, but that soft waters and highly 

 carbonated waters dissolve considerable 

 quantities. 



CoNTiNTjiNG his investigations of the re- 

 cently prepared crystallized calcium, Mois- 

 san describes, in the Comptes Rendus, its 

 action upon nitrogen. In the cold no ac- 

 tion takes place ; at a gentle heat nitrogen 

 is slowly absorbed ; at a low red heat the 

 calcium burns in nitrogen. In these two 

 cases calcium nitrid is formed, of a bronze- 

 yellow color. It is probable that the yellow 

 color previously attributed to metallic cal- 

 cium is due to the presence of more or less 

 calcium nitrid. The calcium of Moissan is 

 a white metal. Calcium nitrid is violently 

 decomposed by water with the formation of 

 ammonia and calcium hydroxid. It reacts 



