April 24, 1884: 



NA TURE 



595 



LETTERS TO THE EDITOR 



[ The Editordocs not hold himself responsible for opinions expressed 

 hy his correspondents. Neither can he undertake to return, 

 or to correspond with the writers of rejected mantiscripts. 

 No notice is taken of anonymous communications. 



[The Editor urgently requests cortespondents to keep their letters 

 as short as possible. The pressure on his space is so great 

 that it is impossible otherwise to insure the appearance even 

 of comii.unications containing interesting and novel facts. ^ 



The Dust of Krakatoa 



In the interesting paper by Mr. John Murray and the Abbe 

 Renard, which appears in your last number (p. 585), there is an 

 erroneous reference which it may be well to correct without 

 delay. I am made responsible for a verbal statement concerning 

 Krakatoa dust which fell in Japan. In your issue of the 3rd 

 inst. (p. 525) a letter from myself will be found, stating, on the 

 authority of Prof. John Milne of Tokio, that, contrai-y to the 

 original statements made on the subject, no dust of Krakatoa is 

 known to have fallen in Japan. My friend M. Renard must 

 have misunderstood the communication which I made to him, 

 which was to the following effect : — I have had the opportunity 

 of examining a great number of specimens of the dust of Kraka- 

 toa which fell at different distances from the volcano, ranging 

 from 50 to nearly 1000 miles. The dust collected at the greatest 

 distance from Krakatoa, with which I am acquainted, is that 

 which fell on board the Arabella in lat. 5° 37' S. and lat. 88° 58' 

 K., Java Head bearing E.A S. about 970 miles. It is certainly 

 true that the dust which has fallen at the greater distances from 

 the volcano contains less magnetite, augile, and hypersthene than 

 that descending nearer to the source of eruption ; and the obvious 

 explanation of this is found in the greater density and compact- 

 ness of the particles of those minerals as compared with the 

 associated glassy fragments. At the same time it must be re- 

 membered tliat this is not the only explanation of the high silica- 

 percentage in these ashes. The prevailing rock in the islands 

 and on the shores of the Sunda Strait ajipears to be a hyper- 

 sthene-augite-andesite, containing an unusually large proportion of 

 a brown, glassy base. This base contains a far higher proportion 

 of silica than the included minerals ; and hence, as shown by 

 Verbeek and Fennema, these rocks have a percentage of silica 

 ranging up to, and even exceeding, 70 per cent. The same is true 

 of the pumices formed from the glassy andesite rocks, including 

 that of Krakatoa itself. John W. Judd 



Hurstleigh, Kew 



On January 13 I collected a sample of snow from an open 

 field, and examined under the microscope the residue left by its 

 evaporation. This residue showed a number of objects which 

 are not usually found in atmospheric dust. Great precautions 

 were taken to prevent the entrance of dust during e\'aporation, 

 the vessel being kept covered with filter-paper. Crystals of 

 common salt were verj' abundant. There were numbers of 

 rather large prismatic crystals, colourless, insoluble in water, 

 and doubly refracting. But the most characteristic objects were 

 minute granules, transparent, colourless, and scattered in thou- 

 sands all over the field of the microscope. These were insoluble 

 in water. Many black particles were visible and some of 

 these were attracted by the magnet. In fact, when the mag- 

 net was swept slowly over the residue, its poles became covered 

 with fine black crystalline particles, evidently magnetic oxide of 

 iron. However, there are large iron-works in this vicinity, 

 which may account for the presence of the magnetic dust. To 

 determine this and other interesting points, it is my intention to 

 examine the snow and rainfall regularly during the next twelve 

 months at least. 



A specimen of snow, freshly fallen on March 10 showed none 

 of the prismatic crystals referred to above. With a high power 

 very small crystals of similar shape and properties were observed. 

 The small granules were, however, to be seen along with crystals 

 of common salt and amnionic nitrate. No magnetic dust was 

 found in this specimen. 



These results are, in my opinion, in favour of the dust theory 

 of the remarkable sunset phenomena of the past winter. 



W. L. Goodwin 



Queen's University, Kingston, Canada, March 31 



P.S. — Snow fell to-day (April i), and a sample was examined 



for dust. The insoluble prisms have completely disappeared, 

 and the minute dust is present in much smaller proportion. 



W. L. G. 



"Earthquakes and Buildings" 



Pi^OF. John Milne, of Tokio, refers in an article under this 

 heading (N.^ture, vol. xxix. p. 290) to buildings in Caracas, iithii k 

 are low, slightly pyramidal, have fat roofs, and are bound alotia- 

 their faces with iron. Beingfor more than twenty years a resident of 

 this city, I hope I may be credited with knowing something of its 

 architecture, and as such I must say that certainly the houses arc- 

 generally one-story buildings, but all the remainder of the fore- 

 going description is quite erroneous. However, I do not wish 

 to make Mr. Milne answerable for its inaccuracies, as it appears 

 to be taken from a ridiculous article published by one Horace I). 

 Warner in the Atlantic Monthly, March 1883. This article is 

 a most audacious fiction from beginning to end, and in none of 

 the statements it pretends to give with graphic seriousness is 

 there any shadow of truth, as I have pointed out in the American 

 Journal of Science, July 1883, with respect to the principal 

 assertion of an earthquake said to have been witnessed by the 

 author on September 7, 1882, in Caracas. 



House-building in our good city is of the most ordinaiy type, 

 and certainly not what it ought to be in a place which already 

 once was ruined by an earthquake (181 2) : the walls are built of 

 brick and mortar ; the roofs are very seldom flat, but have a very 

 slight inclination, say 15 to 20 degrees. They are, however, 

 made too heavy by a thick stratum of loamy mud, spread over 

 the closely-joined laths (generally the stems of the arborescent 

 grass, Ari.ndo saccharoides), on which the tiles are set in alter- 

 nately convex and concave rows. 



The earthquake of Cua (Nature, vol. xviii. p. 130) is an instance 

 of the remarkable influence of the soil on the intensity of de- 

 struction : all the houses built on the rocky lill in the middle of 

 the town W'ere ruined, whilst those on the surrounding alluvial 

 plain suffered scarcely any damage. The same happened in i8l2 

 in Caracas : the northern part of the city, where the stratum of 

 detritus is less deep, was almost completely laid waste ; but the 

 southern part, built on a far deeper deposit of loose matter, 

 experienced comparatively small destruction. A. Ernst 



Caracas, March 16 



On the Transmission of Organic Germs through Cosmical 

 Space by Meteoric Stones 



Ln his addendum to his well-known lecture on " The Origin 

 of the Planetary System " Prof. Helmholti: uses the following 

 remarkable sentence, to which so far as I am aware, attention 

 has not hitherto been directed : — 



" But even those germs which were collected on the surface 

 when they reached the highest and most attenuated kiyer of the 

 atmosphere would long before have been blown away by the 

 powerful draught of air, before the stone reached the denser 

 parts of the gaseous mass, where the compression would be suflS- 

 cient to produce an appreciable heat." 



Helmholtz is contending in favour of the possible transmission 

 of germs from one heavenly body to another, and his point here 

 is that the germs, owing to their being small and light, will be 

 more rapidly retarded (blown back) on reaching the first traces 

 of our atmosphere than the stones on which they reside, and will 

 thus escape the great rise in temperature to which the stones are 

 subject in consequence of friction and air compression. 



Now when a germ just leaves its meteorite its velocity is equal 



to that of the meteorite. If m be the mass of the serm. '"^' 



2/ 

 will be the heat developed in destroying its velocity. Were all 

 this heat to go to raise the temperature of the germ, the rise in 

 temperature would be / = -i^, j- being the thermal capacity of 



the genn. This show^s that the rise in temperature is independent 

 of the mass of the body brought to comparative rest by the atmo- 

 sphere. In reality, since the germ experiences a greater retard- 

 ing acceleration than the stone, its temperature must rise much 

 more rapidly and consequently higher than that of the stone. 

 Further, the terminal velocity of the germ will be less than that 

 of the stone, which will conduce to further raise the temperature 

 of the former. Of course neither the stone nor the germ will get 

 all the heat generated, but this cannot materially affect the 

 question. j. H. Stewart 



Physical Laboratory, Royal College of Science for Ireland 



