SCIENCE. 



949 



Native sulphur is abundant at Mt. Pelee 

 as at many other volcanic vents and results 

 from the reaction of the escaping gases SO, 

 and H.S. The last mentioned gas is readily 

 inflammable and like SO3 and HCl with which 

 it is commonly associated issuing from vol- 

 canoes it is deadly and quickly proves fatal 

 when inhaled in large proportions. To these 

 heavy gases, in part inflammable, most likely 

 commingled with others, is probably due the 

 sudden destruction of life at St. Pierre. 



The above chemical analysis I. by Mr. 

 Steiger shows the presence of .11% sulphur 

 in the insoluble portion of the dust Tests 

 with carbon disulphide indicate that the sul- 

 phur is not free but probably in the form of 

 sulphides. No trace of boracic acid could be 

 found, nor of ammonia or carbonic acid. 

 Salts of ammonia and carbonates are formed 

 only at low temperatures and would fail to 

 leave a record among the solid compounds. 

 Tests for arsenic and antimony were nega- 

 tive also. 



By the kindness of Mr. W. C. Douglas, of 

 the Geological Survey, I obtained a sample 

 of the sand which fell at the Barbados, ninety 

 miles from St. Vincent (one hundred and 

 twenty from Martinique) on the afternoon 

 of May 7. It was collected by Mrs. Mary D. 

 Aughenbaugh, whose interesting observations 

 are published in the Evening Star, Washing- 

 ton, D. C, May 23, p. 7: "Although the vol- 

 canic dust from St. Vincent was coming from 

 the west there was a fairly strong east wind 

 blowing all the time interspersed with hot 

 waves of sulphurous air. The volcanic dust 

 rained continuously here in Barbados from 

 four o'clock P.M. Wednesday, May 7, until 

 Thursday morning at five o'clock and accumu- 

 lated to a depth of three fourths of an inch." 



The particles of sand collected at the Bar- 

 bados are of the same material as those noted 



in the dust collected by Capt. Thomas, al- 

 though differing in proportion, and they evi- 

 dently came from the same source, traveling 

 between Barbados and the Coya in 6 hours, at 

 the rate of nearly 31 miles an hour. Magne- 

 tite appears to be somewhat more abundant 

 and much of it is enclosed in glass. The 

 largest particles have a diameter of about .6 

 mm. with an average of .3 mm., and there- 

 fore a mass of over eight times that of the 

 particles noted above borne to a distance 

 three times as great. The sand from Barba- 

 dos contains a much larger proportion of 

 crystal fragments than the dust from the 

 Coya, for the glassy matter is less than 

 twenty per cent. The dust and sand from 

 St. Vincent drifted mainly to the eastward, 

 for the fall at Kingston, on the southwest side 

 of St. Vincent, as reported by Mrs. Aughen- 

 baugh was about as great as at Barbados, 90 

 miles away. 



When compared with the dust and sand 

 furnished by other volcanoes in recent years 

 it bears the closest analogy with that of the 

 Bogoslov eruption in Behring's Sea, October 

 23, 1883, and collected at Unalaska, sixty 

 miles away. Mineralogically the sands are 

 somewhat different and that at Unalaska is 

 the coarsest, but they are alike in having a 

 decided predominance of crystal fragments 

 over those of volcanic glass. On the other 

 hand, the dust from the great eruption of 

 Krakatoa in the same year wafted many 

 thousands of miles from its source was com- 

 posed chiefly of glass particles, and crystal 

 fragments formed a very small part of the 

 mass. The explosion at Krakatoa was much 

 greater than those of St. Vincent and Mar- 

 tinique. In both cases there was molten rock 

 material erupted by the explosion, although 

 at Krakatoa there was no flow of lava. In 

 Japan, however, a few years ago the only ma- 

 terial erupted was mud, giving evidence of 

 the action of steam without real fusion. The 

 character of the dust and sand examined is 

 such as to indicate that if they were accom- 

 panied by lava streams upon the surface the 

 streams woidd be similar to many flows in the 

 Cascade Range of Oregon instead of the mud 

 flows of Bandaisan in Japan. The similarity 



