460 



NA TURE 



[Ma nil 18, 1886 



minutes of the borrower of the more intelligent class ; to 

 whom, also, the most recent catalogue (and hence the 

 indicator) is generally deficient of three-fourths of the 

 books he most wants, viz. the new ones. 



Nothing is better than the advice given here to secure 

 a good librarian even at a higher cost than some may 

 consider proportionate to the income. But the committee 

 having taken that advice, there is little in this book which 

 will be of value to either him or them. W. Odell 



Les Aerostats dirigeablcs. Par B. de Grilleau. (Paris : 

 Dentu, 1884.) 



This little book does not add anything to the scientific 

 data regarding the direction of balloons which we have 

 lately published ; indeed it was written before the best 

 and most conclusive trials were made. It is a popular 

 view of the subject only ; but it is useful as combating 

 the ignorant prejudice existing thereon in the public 

 mind. It points out to whom the successful solution of 

 the problem is due ; it states the results that have been 

 obtained, and it shows what may be expected to be done 

 in the future. It also explains clearly some of the con- 

 ditions affecting the question, which arc often misunder- 

 stood, such as the effect of the wind, the effective speed 

 obtainable, the nature of the propelling action, and 

 so on. 



LETTERS TO THE EDITOR 



[ The Editordoes not hold himself responsible/or opinions expressed 

 by his correspondents. Neither can he undertake to retttrn^ 

 or to correspond with the writers of, rejected manuscripts. 

 No notice is taken of anonymous communications. 



[ The Editor urgently requests correspondents to keep their letters 

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

 that it is impossible otherwise to iiisure the appearance even 

 of communications containinginteresting and novel facts.^ 



Clifford's "Mathematical Fragments" 

 A SHORT lime since I lent the originals of this work to Mr. 

 A. B. Kempe, F. R.S., as he has been working on the subject 

 of "graphs." Some remarks he made on returning the 

 " Fragments " led me to compare them with the lithographed 

 work, and I propose to supply what is, 1 think, a defect in the 

 publi-^hed book. 



The "Mathematical Fragments" are reproduced on xxii. 

 pages of a uniform size which in the original manuscript U 

 that of the first 5 pages only. The paper of these pages is bhic, 

 and has ruled lines. Page vi. corresponds to two pages of 

 manuscript, indicated by a break, two other pages being blank. 

 Page vii. corresponds to two distinct pages of manuscri|it. 

 Page viii. is made up of three parts, the first six lines on one 

 page of manuscript, the next thirteen of another page of manu- 

 script. These last pages of plain white paper are approximately 

 4I inches wide by 5I inches deep. The page is filled up with a 

 fragment on plain blue paper approximately 5^ by 8 inches. 

 Page ix. is on stout plain white paper, app. 7 inches Ijy 

 9 inches. The last five lines of book are written on the back 

 of the paper. Pages x. -xvl., xviii.-xxi. are written on thin 

 white paper of the same material as page vi., size 9 inches by 

 II inches; the pages are all detached; page 15 is on back of 

 page 14, and page 19 on back of page 18. Page xx. is made up 

 of two distinct pages of manuscript, the second commencing a,t 

 the defaced word which is clearly in the manuscript " Degree." 

 Page xvii. is written in pencil on stiff white paper, folded in 

 half, the lower fragment in the manuscript being in the lower 

 half of the page, and at right angles to the upper — size of full 

 page, app. 7 inches by 10 inches. Page xxii. is made up of 

 three pages of manuscript : the uppermost fragment is on white 

 paper, 6 inches by 4 inches ; the immediately following four 

 lines of writing are given on the back of this page. The next 

 three lines are on paper 7 inches by 10 inches, and the last four 

 lines on precisely similar paper. The time notes at the side have, 

 of course, nothing to do with "graphs." 



These "Fragments" have been circulated (see "Papers," 

 p. 2S6) chiefly amongst libraries ; if the details I here supply 

 are indicated in the copies, their value will, I believe, be greatly 

 increased, and much trouble may be saved by students when 



lliey know how slender a connection there is in some cases 

 between consecutive pages of the text. In the manuscript there 

 is no indication of the order in which the pages should be read 

 beyond what I have pointed out above. The manuscripts are 

 now deposited in the Library of University College, Gower 

 Street. ' R. TUCKER 



The Upper Wind Currents in the South Indian Ocean 

 and over the N.W. Monsoon 



AiTER sending a brief account to Nature of my observations 

 on the upper wind currents over tlie Atlantic doldrums, I started 

 from Natal for some journeys across various portions of the 

 Indian Ocean, to investigate the circulation of tlie higher atmo- 

 sphere in that region. 



I first went to Mauritius. During the whole passage from 

 Natal, at the end of December, we sailed in the S.E. Trade, 

 with an almost constant movement of high cirrus from the N.W. 



Though I was disappointed in not meeting with a cyclone in 

 those seas, still I succeeded in obtaining much valuable informa- 

 tion about the details of hurricane weather, which could only be 

 learnt on the spot. One point relating to upper currents is very 

 important. The cirrus which appears five or six days before the 

 arrival of a hurricane follows the normal course from N.W. or 

 .S.W. , and is no guide to the path of the cyclone. But on the 

 outskirts of the hurricane, low clouds afford valuable informa- 

 tion. If the cloud over the S.E. surface-wind inclines towards 

 E. the centre of the cyclone will pass to the N. ; if on the con- 

 trary the low cloud inclines towards S. the centre will pass to 

 the S. of the observer. 



Though Meldrum, and Bridet'of Reunion, both agree on this 

 point, the subject requires further elucidation, for such a rota- 

 tion of upper currents is contrary to all analogy of what is sup- 

 posed to hold round cyclones in the northern temperate 

 regions. I am certain from my own investigations that the 

 general character of tropical and extra-tropical cyclones is 

 identical. In Mauritius hurricanes I find the same oval form, 

 the same squall at the turn of the bai'ometer, the same halo in 

 front, and hard, detached cloud in rear, which characterise 

 European cyclones. Mr. Harris has recently traced a cyclone 

 from its easterly course as a typhoon in the China Seas, across 

 the Pacific, United States, and Atlantic into Western Europe. 

 Like eveiy other long-lived cyclone, this one received accessions 

 of strength by fusion or coalescence with others which had 

 formed outside the tropics. It is perfectly certain that cyclones 

 which revolved on difi'erent systems could not unite, and I think 

 that the motion of the lower layers of cloud over the northern 

 side of our own cyclones should receive special attention. At 

 present we are led to believe that the cirrus in front of a cyclone, 

 both right and left of the path, comes from S.W. or S. 



Be this as it may, cloud motion forms a useful adjunct to a 

 valuable and successful system of hurricane forecasting that is 

 carried out by Mr. Meldrum, who, in the absence of telegraphs, 

 has to rely entirely on his own instruments and above all on his 

 own experience and judgment. Another interesting feature of 

 this system is the care which must be taken to allow for the 

 diurnal motion of the barometer during the slow diminution of 

 pressure which always precedes tire arrival of a hurricane. 



From Mauritius I sailed to Adelaide, so as to examine the 

 Polar limit of the S.E. Trade. Though we steered a great 

 circle course which took us fully into 39° S. latitude, we ex- 

 perienced constant S.E. and E. winds. These must have been 

 due to some extra-tropical anticyclone, and every observation of 

 low or middle layers of cloud showed a current practically in 

 the same direction as the surface-wind. 



At Adelaide I learnt that the normal direction of the highest 

 currents is from N.W. In that city the direction of the surface- 

 wind is much influenced by land and sea breezes. Through the 

 kindness of the actinig chief of the Observatory, Mr. W. E. 

 Cooke, I was able to confirm what I had previously suspected 

 from my investigations in Melbourne last year, that sometimes 

 at least the characteristic "southerly bursters" of Australia, 

 are due to that class of V-shaped depression in which the rain 

 occurs in rear of the disturbance. Other times the sudden irrup- 

 tion of S. wind appears due to the shift of wind at the passage 

 of the trough of a cyclone. 



From Adelaide I came here to Colombo so as to repeat a 

 section of Indian Ocean very nearly at the same season and in 

 the same straight line as last year. The results of the former 

 voyage were communicated to Nature, vol. xxxii. p. 624, 

 when I announced the fact that the highest currents over the 



