June 2i, 1888] 



NA TURE 



i«i 



subsequent consideration at all, we may suggest a very 

 different direction in which they might be modified in actual 

 fact, and in which their spirit would yet be even better 

 represented than by a literal fulfilment. It was pointed 

 out that in taking the photographic plates of stars down 

 to the 14th magnitude in parts of the sky where brighter 

 stars existed, these with the exposure necessary to obtain 

 the 14th magnitude would be very much over-exposed. 

 And it was suggested that it would be advisable to take a 

 second series of plates, as already mentioned (see Resolu- 

 tion 17). Now in some parts of the sky no second series 

 of plates are, from this point of view, at all necessary ; 

 whilst in others not one or two, but many series of plates 

 would be necessary in order to do justice to the various 

 magnitudes in that particular part of the sky. For the 

 present this is not the point at issue, but it may serve as an 

 illustration of the sort of interpretation of the resolutions 

 which we should consider legitimate. 



In order to come to a proper judgment on the legiti- 

 macy of the derivation of Dr. Gill's proposal from the 

 resolutions it is necessary to make some statements, which 

 are not new, but of which the true significance does not 

 seem to have been universally appreciated : — (1) When 

 the plates are obtained they are actual representations of 

 the stars as existing at a given time, and for every purpose 

 except spectrum analysis are as good, if not better, than 

 the visible heavens. If with these plates we have the 

 absolute places of a certain small number of known stars, 

 we have then all the data to make them valuable, either 

 in the present or in the future. (2) The many questions 

 concerning the stars which it is hoped a photographic 

 chart of the heavens would do a great deal towards 

 settling, such as their distribution, their proper motions, 

 their changes of magnitude, and the presence of minor 

 planets, of new stars and the like, can all be best treated 

 by a direct comparison of plate with plate, in any of the 

 various ways in which this can be done. (3) In order to 

 obtain the best results from such an agent as photography 

 it is necessary to use it in its own proper way ; and 

 astronomers must recollect that old methods of procedure 

 adapted to other instrumental means may most probably 

 be out of place. We might considerably enlarge on these 

 statements, but for our present purpose it is sufficient to 

 call attention to them. 



Now, if Dr. Gill's catalogue were successfully con- 

 structed — and there are, alas ! many difficulties in the 

 way — its utility in the direction of comparison of our sky 

 with that of the future is wholly limited by one condition, 

 that in the future another exactly similar catalogue be 

 constructed, occupying a similar time. Even then, if any 

 changes were found by means of this comparison of 

 catalogues which might very well be made in the course 

 of fifty or one hundred years, the natural and indeed the 

 proper thing to do would be to immediately compare the 

 original plates. But can it be possible that any man or 

 number of men really think of dealing with such a subject 

 in such a way ? If, on the other hand, the object of a 

 catalogue be merely to allow of comets, minor planets, 

 and other bodies being located, surely it would be better 

 to measure the plates as occasion arises, and not to cata- 

 logue 2,000,000 stars on the off-chance of having some 

 twenty or thirty positions to settle in the course of a year. 

 And, further, such a catalogue would have this enormous 

 disadvantage, that whilst in some parts of the sky stars 

 of the nth magnitude would be fairly well spread, in 

 the Milky Way we should have stars clustered in such 

 enormous quantities that it would be an extremely difficult 

 thing to even identify them : in fact, speaking roundly, 

 we should say that if such a catalogue were made, two- 

 thirds of the stars catalogued would lie in the Milky Way. 

 If, contrary to the opinion we have expressed, it is 

 decided to form a very large catalogue, surely it would be 

 better to determine the places of a certain number of stars, 

 of such magnitudes as are found available, in each square 



degree, and make these the reference stars from which 

 the positions of the other stars on the plate could be 

 obtained. 



We are therefore of opinion that, supposing limitless 

 time and money available for such a purpose, the advant- 

 ages of constructing this catalogue would be doubtful ; 

 but even if we waived all these objections and agreed that 

 such a catalogue would be a " nice thing to have," or 

 admitted that since men of the ability and reputation of 

 Admiral Mouchez and Dr. Gill consider such a catalogue 

 necessary it is heresy to inquire the why and wherefore, 

 there would still be left the serious objection that to form a 

 chart of the heavens is the first thing to do, and, take it 

 in as simple a form as possible, it will quite possibly tax 

 the energies of astronomers to their utmost ; and that 

 stellar photography being as yet in its infancy it is suicidal 

 to attempt anything which will commit us to a course of 

 action extending over more than a very few years. We 

 could not give a better illustration of the dangers of the 

 opposite procedure than has been supplied by Admiral 

 Mouchez himself. In a recent article he has suggested 

 that there have lately been such improvements in the 

 sensitiveness of plates that we could now go to the 15th 

 magnitude instead of the 14th. With a little ingenuity 

 and less arithmetic it could easily be shown that the 

 whole plan of operations would have become hopelessly 

 futile and obsolete before half the time allowed by Dr. 

 Gill for its completion had elapsed. 



But not for one moment do we wish to appear lacking 

 in sympathy with those who have spent and are spending 

 so much time and thought on this subject ; it is our great 

 anxiety for the success of the work in which they are co- 

 operating which makes us eager to protest as far as we 

 can against the grand mistake of attempting too much. 



THE INCURVATURE OF THE WINDS IN 

 TROPICAL CYCLONES. 



THE question of the incurvature of the winds in tropical 

 cyclones is one of such importance to mariners, to 

 enable them to judge their position in a storm, and to 

 escape the hurricane around the central calm, that no 

 apology is needed for adding my independent testimony 

 to that of Prof. Loomis, whose conclusions, given at length 

 in his recent well-known memoir, " Contributions to 

 Meteorology," are quoted in Mr. Douglas Archibald's 

 paper on M. Faye's work " Sur les Tempetes " in. last 

 week's Nature (p. 149). 



In the preparation of a forthcoming work on the weather 

 and climates of India and the storms of Indian seas, I have 

 lately had occasion to re-investigate the above question 

 in the case of cyclones in the Bay of Bengal, on the 

 evidence afforded by the numerous original memoirs and 

 reports prepared by Messrs. Willson, Eliot, Pedler, and 

 other officers of the Indian and Bengal Meteorological 

 Departments ; my object being the practical one of de- 

 termining directly the bearing of the storm-centre from a 

 ship's position ; and instead, therefore, of measuring the 

 angle between the wind direction and the nearest isobar, 

 as was done by Prof. Loomis, I have measured with a 

 protractor the angle included between the former and 

 its radius vector, in all cases in which the position 

 of the storm's centre has been ascertained on sufficient 

 evidence. In one other important condition I have also 

 departed from the method pursued by Prof. Loomis. I 

 have restricted the measurements to wind observations of 

 ships at sea, within the influence of the storm, and to 

 those of good observatories on the coast, subject to the 

 same proviso ; and have taken no account of those of 

 inland observatories. This difference of procedure is 

 probably the reason that the amount of the incurvature 

 shown by these measurements is somewhat different from 



