May 25, 1893] 



NA TURE 



77 



light of Japan" had, as was customary at this season of the year 

 when the weather is very cold, stormy, and clear, been observed 

 by fishermen in the Shimbara Gulf and Japanese waters. The 

 article went on to say that these lights were referred to in native 

 school-books, and attributed to electrical phenomena. On 

 mentioning the matter, however, to the leading Europeans in 

 Yokohama and Tokio, they appeared to have no knowledge of 

 the matter. 



Captain Castle, of H.M.S. ZeaW^r, informed me that, not 

 long ago, the officers of his ship saw lights in the same locality 

 which they thought at first were caused by a ship on fire. The 

 course of the vessel was altered at once with a view of rendering 

 assistance, but finding that the lights increased their altitude as 

 he approached, he attributed them to some volcanic disturbance, 

 and being pressed for time, resumed his course. 



The background of high land seen on the first night dispels 

 all idea of these extraordinary lights being due to a distant 

 volcano. The uniformity of the bearing renders the theory of 

 their being fires on the shore most improbable. I am inclined 

 to the belief that they were something in the nature of St. Elmo's 

 fires. It is probable that there are travellers among the readers 

 of your interesting journal who have seen or heard of this 

 phenomenon, and will be able to describe its origin and the 

 atmospheric conditions necessary for its appearance. 



Chas. J. NoRcocK. 



H.M.S. Caroline, Hongkong, April 10. 



The Greatest Rainfall in Twenty-four Hours. 



I\ Naturk, May 4, Mr. Clement Wragge, of Brisbane, 

 confidently asserts that Queensland has beaten the world's 

 record in the extraordinary amount recorded on P'ebruary 3, 

 viz., 357 inches. I am sorry tt) have to take away such an 

 unenviable palm from Queensland, by recalling a fact well 

 known to every Indian meteorologist that the highest record 

 extant belongs to Chirapunji, in the Khasia hills, where on 

 June 14, 1876, 40'8 inches were recorded in the twenty-four 

 hours. Not only so, but on the 12th 30 inches fell, and in the 

 four days, from the I2th to the 15th inclusive, as much as 102 

 inches. Of course the effects were not so disastrous in this case, 

 as indeed such a state of things is little removed from the normal 

 at Chira in the early part of June, but I have a very clear recol- 

 lection of it as I was at Chirapunji on the 12th and 13th, and 

 not far from it on. the memorable 14th. 



The conditions which have occurred in Queensland and the 

 North Island of New Zealand during the last six months have 

 been a remarkable example of persistent abnormals, and though 

 the total number of rational causes may still be wanting to 

 explain everything, one or two were evidently in operation 

 when I was there from October to January, and I am confident 

 that from the empirical law of persistency, coupled with a few 

 rational inferences, a forecast of impending floods could have 

 been made and can be made for the future, much in the same 

 way as the general character of the monsoon can be foretold 

 in India. 



May 13. E. Douglas Archibald. 



A Dust-whirl or (?) Tornado. 



Ix Nature (vol. xl. p. 174) you kindly allowed me to 

 describe a dust- whirl seen to originate on a heated dust-covered 

 highway. The phenomenon has just been repeated under much 

 similar circumstances, only in this instance the column of dust 

 after oscillating to and fro on the highway for about half a minute, 

 moved rapidly away in a curvilinear path in a northerly direc- 

 tion, the lower end of the whirl catching up loose material in its 

 track where it touched the ground, which it did at intervals of 

 from ten to fifteen yards, carrying the strawv litter from a straw- 

 berry bed upwards of 50 yards in the air. It appeared to dissi- 

 pate into the upper air when crossing a meadow some 300 yards 

 from its place of origin. The characteristic "swish "of the 

 rushing air was very marked, and the four motions common to 

 all tornadoes (see I.ieut. Finley's "Character of Six Hundred 

 Tornadoes"), viz. whirling from right to left, progressive motion 

 to the north, a curvilinear track, and the dipping up and down, 

 were all distinctly traced. The question therefore, naturally 

 arises — Can these dust-whirls be tornadoes in miniature? 



Conditions at the time of the occurrence : — D,>te, Thursday, 

 May II, 1893; time, II a.m. Corrected barometer, 30.327 

 (falling slightly). Dry bulb, 65°.5 ; wet, 5l".8 = rel. hum. 

 38 per cent. Wind, south ; force, I. Some upper cirrus 

 radiating from north-east, and drifting slowly from north-west, 



NO. 1230, VOL. 48] 



showing top and bottom arcs of halo at 10 a.m. Black bulb 

 it: vacuo 128°. 2 ; weather very warm and dry. , 

 Driffield, May II. J. LovEL. 



What becomes of the Aphis in the 'Winter ? 



I have spent many weeks this spring closely observing the 

 budding trees, with the object of discovering in what condition 

 of life the aphis spends the winter ; as the result of my obser- 

 vations, which were made under the microscope, I believe that 

 the aphidte during the autumn (or as many of them as have 

 reached the state of reproduction) attach themselves to the stem 

 of the tree, with their young inside them, in much the same way 

 as the female members of the closely-allied family coccid^ 

 do. In course of tirre the mother-aphis becomes simply a dried 

 skin serving as a p' jtection to the yourg. When the warm days 

 of spring come these are developed and easily make their way 

 through the skin and crawl on to the young leaves, there to 

 begin their work of sucking and reproduction. 



T. A. Siiarpe. 



Soot-figures on Ceilings. 



May I suggest a distinct, if not an alternative cause for 

 Prof. E. B. Poulton's soot figures in Nature, April 27th? 

 The ceiling plaster is very porous, except where it is in contact 

 with the joists, etc. At such points veiy little deposit occurs 

 compared with the spaces where the hot air is vigorously diffus- 

 ing through into the cold space above. I suggest this because 

 I am very familiar with a large ceiling where the rafters are 

 thus picked out in light shades. Even the laths are picked out, 

 but less distinctly. The main bolts likewise show dark, as in 

 Prof. Poulton's sketch, as if there were an air-space by them. 

 There is no perceptible difference in the figures near the central 

 chandelier from those in the corners remotest from heating 

 causes. The bombarding pattern is often very well shown 

 where super-heated water pipes rvin along a white-washed wall. 

 The effect of every little break, even a nail in the wall, is most 

 striking. J. Edmuxd Clark. 



A Difficulty in Weismannism Resolved. 



In' my letter of the 1st inst. an omission of parentheses and 

 quotation marks, which I omitted to note on the proof, alters 

 the sense of the paragraph with quotations from the "Germ 

 Plasm," pp. 434-5. It should be as follows : — " The note runs 

 thus : ' Compare Marcus Harlog, Nature, vol. xliv. p. 102,' 

 (the reference omits my letter of Oct. 31, 1891). 'The deductions 

 made by this author are logically correct but are no longer justi- 

 fiable,since I myself have gained further insight into the problems 

 concerned.' " The absence of the inverted commas disguises this 

 recognition by Weismann of the validity of my objections, and 

 of the consequent change in his own views. 



Cork, May 15. Marcus Hartog. 



NOTES. 



The Hon. Ralph Abercromby has given to the Royal Society 

 of New South Wales the sum of;^ioo, which is to be offered 

 as prizes with the object of bringing about exhaustive studies of 

 certain features of Australian weather. So far only one feature 

 has been selected, and a prize of £2$ is now offered for an ex- 

 haustive study of the well-known " Southerly Burster." Itisun. 

 derstood that no essay which does not deal fully with the follow- 

 ing points will be considered : — (l) The motions of the various 

 strata of clouds for some hours preceding, at the time of, and 

 following the "burster;" (2) the weather conditions which 

 lead up to and follow the "burster," with weather charts of 

 Austr,ilia for the day of occurrence and the following day ; (3) the 

 general conditions which modify the character of the " burster ; " 

 (4) The area of the "burster" and its track ; (5) barograph traces 

 showing the changes of pressure during the "burster;" (6) 

 the direction and character of wind preceding it ; (7) the relation 

 of " bursters " to rainfall. The essay must not exceed 50 pages 

 of foolscap, and must be sent in not later than March 31, 1S94. 

 It must embody studies of several "bursters," and must be 

 chiefly the result of original research of the author, but authors 



