OBSERVATIONS IN CHEMISTRY AND MINERALOGY 417 



at red heat. Duration of heat in hydrogen 1 hour. Loss of weight 0.0055. gr 

 The tube shows a pale ruby color very uniformly spread exocpt when asbestos 

 touched. 



XV III. Same bundle in new tube, oxidized in horizontal position at red heat 



during 20 minutes; reduced in hydrogen during 1 hour 20 minutes. loss 

 0.0005. Tube shows no color in transmitted light, but pale copper red in reflected 

 light from dust copper. Noticed that asbestos cord nearest to copper shows ruby 

 color; first instance of the kind. 



XIX. Same bundle in fresh tube, allowed to oxidize during 35 minutes * 

 red heat. Reducing period in hydrogen 1 hour. Loos of v ight is 0.0013 

 Development of ruby, stroruj and even, 



XX, XXI, XXII reveal nothing special and hardly a trace of coloration. 

 XXII I. Same bundle. Oxidation as above during 45 minutes. Deoxidation 



during l}/£ hours. Loss in weight is 0.0003 grams, development of ruby is mag- 

 nificent between the cords, that is between A and B. Forward of A the color 

 is very pale, while backward of B the ruby is stronger and f!io\ besides two 

 rings of dust copper, shown in figure with dots. Temperature w:is higher at A 

 than at B. The strongest development is downward. The gas-current proceeds 



with the arrow (fig. 10, PI. XXXVI). 



Experiments XXIV to XXXV were made chiefly in order to find the most 

 favorable conditions for oxidation, so that the film should not peel off. Af t er each 

 oxidation, however, the deoxidation was carried out without giving a dec led 

 ruby color in any of the ten cases. At the end of description of XXXIV my 

 notebook says: "The mystery of the formation of the ruby is now more bewil- 

 dering than ever before." The glass showed a brown smokiness, very pale, 

 in the places where the ruby usually appears. Regarding the avoidance of 

 peeling the finding pointed to a slow action at a temperature of 490° to 520* C. 

 The placing of the tube horizontally with both ends partly or entirely open gives 

 a slow exchange of the nitrogen inside with the air outside. 



XXXVI. The material in this experiment comes from the Michigan Mine. 

 The wire was drawn from ingots which I had melted myself from the chips of 



mass 



The material contains Cu - 99.8199: Ag - 0.1025; Fc - 0.00 



As = 0.0060; O = 0.0066. It is therefore exceptionally pure copper. Its 

 electric conductivity is 100.7. The wire was rubbed off with chamois leather, 

 but not burnished and scraped. The bundle was oxidized slowly during 5 hours 

 and deoxidized during 2 hours at temp. 690°-710° C. The display of ruby wa$ 

 great, as seen in fig. 11, pi. XXXVI. 



Drawing is one-half natural size. The current is with the arrow. 



A purple-colored ring surrounded by transparent ruby. 



(6) Without color because protected by asbestos. 



(c) Almost uniform mixture of ruby and dust-copper. 



(d) Without color because protected by asbestos 



(e) Mixture of ruby and dust-copper. 



27 JOURN. ACAD. NAT. SCI. PHILA..VOL XV 



