19 



able. Favouring the condensation of a hazy vapour it would, 

 however, primarily give rise to the sun-glow. I shall presently 

 endeavour to show why the dust does not fall quickly to the 

 €arth, no matter what may be its origin. Possibly it may be 

 related to the cause of the zodiacal light, which, according to 

 one hypothesis, is formed of myriads of solid particles. 



Coming now to the telluric hypothesis, the mass of evidence 

 relating to the Javan eruptions points strongly to Krakatoa 

 as the cause. The eruptions commenced in May last, cul- 

 minating in the fearful convulsions of August 26th. Consider 

 the immense amount of volcanic dust erupted ; counted probably 

 by millions of tons of dust, and exceeding by far the dust con- 

 tributed by meteorites during the whole year. Consider the 

 tearing asunder of the atmosphere over that mighty furnace, 

 and the height to which particles of dust would be launched. 

 Some idea of the atmospheric wave generated is gathered from 

 the fact that the barometer on board a steamer 150 miles distant 

 rose and fell half-an-inch every two or three minutes, as re- 

 ported by Mr. Bishop, of Honolulu. Observations show that 

 this gigantic air wave was visible on the barometer curves at 

 several stations for five days after its origin ; that its velocity 

 was 674 miles an hour, that it travelled before its extinction, 

 more than 82,200 miles, and that it passed Si times round the 

 -entire circumference of the earth. ("Nature," No. 738, 

 p. 182). Apart from the blood-red and green suns, showers of 

 dust and pumice and other startling phenomena, of which we 

 have heard so much, and which I pass by, my opinion as to 

 Krakatoa being the cause is greatly strengthened by the over- 

 w^helming fact that material brought down by rain in Holland 

 and snow in Spain — crystals of hypersthene, pyroxene, and 

 magnetic iron with volcanic glass — has been proved to be iden- 

 tical with the matter found in the Java ashes. It is objected, 

 with much reason, to this volcanic theory : — How could the 

 dust have spread so rapidly as to reach Trinidad in a 

 week, while it took double that time to travel to India and 

 Ceylon. This objection seems as fatal to the volcanic hy- 

 pothesis as does Mr. Ellery's to the meteoric one ; and with 

 our knowledge of the general distribution of atmospheric 

 pressure at that season of the year, is a difficult one to meet. 

 Our first thoughts are of the great low pressure areas of 

 Central Asia at that summer season ; and we wonder why the 

 heavier air currents which set inland from over the cooler 

 ocean to take the place of the heated ascending currents of the 

 Asiatic continent, did not convey the dust to India first. Perhaps 

 closer consideration will enable us to give a satisfactory answer. 

 The eruption of Krakatoa took place in a zone of low barometer, 

 in a branch of the equatorial calm belt. My own observations on 



