152 



SCIENCE. 



[N. S. Voii. XVII. No. 421. 



or orange, on horizon, purple glow getting 

 somewhat brighter; 4:30 p.m., orange glow 

 continues on horizon, the afterglow has 

 changed from purple to pink and is much 

 more brilliant, extending to within 7° of the 

 horizon, maximum brilliancy about 15° alti- 

 tude; 4:35 P.M., the orange glow continues 

 on the horizon, maximum brightness of the 

 afterglow about 10° ; 4 :40 p.m., orange glow on 

 horizon growing fainter, the maximum bright- 

 ness of the secondary glow is at an altitude of 

 7° and is growing somewhat fainter; 4:45 p.m., 

 primary joined to secondary glow and forms 

 a bright orange band about 6° wide on horizon; 

 4:50 P.M., bright red band about 3° wide on 

 horizon, with here and there short streams 

 extending toward zenith, a faint purple ter- 

 tiary glow has appeared at an altitude of about 

 45° ; 4 :55 p.m., red glow about 3° wide on 

 horizon, tertiary glow brighter and extending 

 from altitude 20° to zenith, maximum about 

 40° ; 5 :00 p.m., tertiary glow bright, with maxi- 

 mum about 20° ; 5 :05 p.m., red glow on horizon 

 fading, purple tertiary, glow still bright, with 

 maximum about 15° ; 5 :10 p.m., tertiary glow 

 fading, maximum about 10°; 5:10 p.m., red- 

 dish glow about 5° wide on horizon lasting 

 until 5:25 p.m., when it began to fade rapidly; 

 5 :30 P.M., red band on horizon about 1° broad 

 and growing faint; 5:35 p.m., reddish glow 

 still visible; 5:40 p.m., glow gone. 



The duration of these sunsets was consid- 

 erably longer than the normal sunsets, and 

 it is probable that they were due to the dust 

 from the West Indian volcanoes. 



Henry Helm Clayton. 



Blue Hill Meteorological Observatory. 

 January 11, 1903. 



SHORTER ARTICLES. 

 some corrosions found on ancient bronzes. 

 At the suggestion of Gen. C. G. Loring, 

 of the Boston Museum of Fine Arts, I under- 

 took, some years ago, the investigation of va- 

 rious corrosions appearing on ancient bronzes. 

 The large collection of Grecian and Egyptian 

 bronzes in the Boston Museum furnishes suf- 

 ficient variety to make the observations of 

 general value. The results obtained may, 



therefore, be of interest to collectors and 

 curators of other museums. 



The ordinary dark green corrosion or 

 patina familiar to every one, and most com- 

 monly observed on bronze statues exposed to 

 the weather, consists of basic copper carbon- 

 ate and is comparatively harmless. On very 

 old statues, especially if they have been buried, 

 two different corrosions have been noticed, 

 which from their appearance may be desig- 

 nated as the pale blue and pale green excres- 

 cences. As will be seen later, both may en- 

 danger the life of the bronze, and especially 

 the pale blue is the seat of an active chemical 

 reaction. 



The pale Hue excrescence occurs in blotches 

 all over the surface of the bronze and is espe- 

 cially noticeable in less exposed parts, such 

 as indentations or cavities. It has a very fine 

 powdery appearance and, on account of the 

 ease with which it seems to spread from one 

 bronze to the other, was supposed to be of bac- 

 terial origin.* A chemical analysis, however, 

 indicates a different origin of the corrosion. 

 About a gram of the substance was carefully 

 collected and found to consist of 50 per cent, 

 sodium carbonate, 25 per cent, copper carbon- 

 ate, 25 per cent, sand and a trace of sodium 

 stannate. The large percentage of sodium 

 carbonate leads to the following theory as to 

 the origin of the blue rust: As long as the 

 bronze lay buried in the dry Egyptian soil, 

 no reaction took place ; on exposure to a moist 

 atmosphere, however, some moisture con- 

 densing gave the carbonic acid of the air a 

 chance to combine with the sodium carbonate, 

 forming acid sodium carbonate. This then 

 attacks the metal, forming copper carbonate 

 and regenerating the sodium carbonate, which 

 combines with the copper carbonate to form 

 a double salt, thus accounting for the blue 

 color. It is easy to see how in the course of 

 time a large amount of metal may be thus 

 corroded. To test the above hypothesis a 

 fresh piece of bronze and some powdered 

 sodium carbonate were exposed for several 

 months to a warm, moist atmosphere. The 

 pale blue excrescence appeared and was iden- 

 tical in all respects with the original rust col- 

 * Dr. Wm. Frazer, Nature, 1898, May 19. 



