672 



NA TURE 



[October 31, 1907 



aieai-ly the same time that the latter completes igi revolu- 

 tions; the periods are i6,8oi and 1(1,802 days respectively, 

 and thus transits of Mercury must occur every forty-six 

 years at the same node. 



TiiK Perseid Meteors. — In No. 4206 of the Asttono- 

 mische Nachrichtcn (p. 81, October 16) Hcrr \V. Milo- 

 wanov, of the Kasan University Observatory, Russia, gives 

 an account of the Perseid observations made at Kasan in 

 .August, 190b. Two hundred and seventy-two meteor-paths 

 were recorded, by three sets of observers, on August 11, 

 12, and 13, forty-nine of which were not Perseids. The 

 horary rate on .\ugust 11 was 21-5, and on .August 12 

 25-4. The time of observation, the path and the magni- 

 tude of each meteor are given, together with a list of the 

 variously estimated radiants ; the centre of the radiant for 

 iqo6 is given as a = 43°.5, 5=4-55"'o. The heights of 

 twenty-four meteors are also given, the mean heights of 

 «ippearance and disappearance being iii km. (sixty-nine 

 miles) and 73 lim. (forty-five miles) respectively. 



A Rich Nebula Region. — On a plate exposed in the 

 Kruce telescope for three hours on July 16, igo6, Prof. 

 Max Wolf found an extensive nebulous cloud having its 

 centre some i| degrees north-east of e, Sagittarii. Later 

 observations show that the region is very rich in small 

 nebuUc, a plate exposed on July 16 showing a large number 

 of such objects over a region of forty-eight square degrees. 

 Most of these nebula; are of Prof. Wolf's class I,, being 

 round objects with central condensations (Asironomische 

 Kachrichteti, No. 4207, p. 109). 



RAIN-GAUGE EXPOSURE AND PROTECTION. 

 AXTHEN during the nineteenth century rainfall observ- 

 ations by means ot gauges began to be carried out 

 upon an extensive and scientific system, it was soon dis- 

 covered that gauges which stood in situations much ex- 

 posed to the wind invariably indicated a smaller amount of 

 rain than such as were protected from the wind. 



The greater part of our knowledge regarding the effects 

 \)f wind exposure upon the indications of rain-gauges 

 was derived from a careful study of the diminution of 

 recorded rainfall with the elevation of gauges, cither upon 

 buildings or poles, above the ground, and an abundance 

 of experimental work carried out during the nineteenth 

 century in many European countries, notably, perhaps, in 

 England, as well as in the United States of .America, left 

 no doubt but that this indicated decrease of rainfall with 

 height above the ground was only apparent, and due to 

 the more imperfect catch of rain by the gauges consequent 

 upon their increased exposure to wind. 



The well-founded conclusion is that wind interferes with 

 the proper catch of rain by eddying around the mouth of 

 the rain-gauge, and that consequently a rain-gauge should 

 on theoretical principles be protected from wind disturb- 

 ance if we w-ould know the true quantity of rain that 

 falls upon the ground in its vicinity. Nothing is known 

 as to the mode of formation or of the complexity of these 

 wind eddies, and information respecting these questions 

 could only be suggested through laboratory experiments. 

 When a gauge is elevated above the ground upon a thin 

 pole there is merely the augmented wind velocity to be 

 considered in explanation of the decreased amount of rain 

 it will receive as compared with a similar one upon the 

 ground, but when a gauge is placed upon an edifice or 

 close to the edge of a steep cliff or bank the case is far 

 more complicated through new disturbances introduced by 

 such obstructions themselves to the wind, which has the 

 effect of causing a deficit of rain upon the windward side 

 of buildings and a relative surplus upon the leeward side. 



The whole subject, however, of decrease of recorded 

 rainfall with height above the ground has been thoroughly 

 Investigated, as mentioned above, in its several relations, 

 so that there offer themselves for discussion, more par- 

 ticularly the methods that have been adopted for protect- 

 ing rain-gauges from, or of correcting their readings for, 

 w'ind error. 



It was about the year 1889 that Prof. Cleveland .Abbe,' 

 of the United States, made an exhaustive study of the data 



J- '^^."J'l'y W'-nther Re^m-.K. vol. xxii., 1894, p 25. "The Reliability of 

 *■* " " "" ' ' <in Meteorological Journal^\o\,\\. ^. i^iy^^'Va^ 



NO. 1983, VOL. 76] 



at that time available from different parts of the world 

 with reference to the apparent decrease of rainfall with 

 elevation above ground, and relating his data to a law 

 known to meteorologists as .Archibald's, connecting in- 

 crease of wind velocity with square root of altitude for 

 small altitudes above the surface of the earth, showed 

 that the deficit of rain indicated by an elevated gauge was 

 proportional to the square root of its altitude above the 

 ground. Erom these results .Abbe deduced for a rain- 

 gauge in a free, open situation a numerical wind correction 

 which may be explained as follows : — 



If a second gauge, in all essential respects similar to 

 the one for the readings of which the correction is to be 

 applied, be placed twice as high above the ground as the 

 latter, the quantity EX2-4i4 + R = C, where E is the excess 

 of the reading of the lower gauge above that of the 

 upper, R is the reading of the lower gauge, and C is the 

 corrected reading of the lower gauge ; to state this in 

 words, add to the reading of the lower gauge 2414 times 

 its excess above that of the upper, and the result is the 

 amount of rain which the lower gauge would have caught 

 in the absence of wind disturbances. The validity of this 

 correction, which depends, of course, upon the extent to 

 which for any particular occurrence of wind and rain the 

 necessary assumption of .Archibald's law is trustworthy, is 

 doubtless such as to enable at least a close approximation 

 to the true rainfall to be reached. But two conditions 

 must be complied with before the records of two gauges 

 placed .at different heights above the ground can thus bear 

 a simple relation to differences of wind velocity corre- 

 sponding to different amounts of error through eddy action 

 at the mouths of the gauges : — (a) the two gauges must 

 be of the same form and dimensions ; (b) if the lower 

 gauge rests upon the ground, the upper one must be sup- 

 ported on a somewhat thin pole as may not induce, by 

 offering an obstruction to the wind, other disturbances 

 operating at the mouth of the rain-gauge in addition to 

 those due to the gauge itself. 



Of the various types of wind-shield for rain-gauges that 

 have been suggested or tried, by far the most serviceable 

 is the protecting funnel jacket ' originally devised about 

 the year 1879 by Prof. K. E. Nipher, of the Linited States. 

 This is screwed on to the cylinder of the rain-gauge at 

 such a height that the rim of its broad upper portion lies 

 level with that of the mouth of the gauge, and consists 

 usually of copper gauze for the purpose of preventing or 

 minimising the splashing of water into the gauge from 

 the inner surface. The efficiency of such a protector has 

 been tested, not only by Nipher himself, hut by Prof. R. 

 Bornstejn in Germany and Dr. H. Wild in Russia, with 

 the result that the contrivance may, on the whole, be re- 

 garded as an appropriate means of eliminating, or at all 

 events reducing, injurious wind eddies around the mouth 

 of a rain-gauge, as will be shortly seen. Since the pro- 

 tecting jacket was invented by Nipher it has been improved 

 in various ways, a comparatively recent modification, as 

 used at some of the Russian rainfall stations under the 

 direction of the Central Physical Observatory at St. Peters- 

 burg, being capable of being taken to pieces so as to 

 facilitate packing for transport. 



.Another type of protection contrivance, more accurate 

 than the Nipher jacket, was established about 1880 in 

 St. Petersburg by Dr. H. Wild. This took the form of 

 a rectangular fence enclosure composed of wooden slabs, 

 2-5 metres in height, at certain equal distances on each 

 side from the rain-gauge, the mouth of which was i metre 

 above the ground, and provided with a door for the entry 

 or exit of the observer. \ comparison of anemometer 

 observations within and without such an enclosure shows 

 that even during the prevalence of high wind the con- 

 dilions within approximate nearly to those of a calm, so 

 that a fence enclosure of the dimensions adopted by Wild 

 may be taken as the standard of accuracy for wind- 

 protection contrivances. 



' Zeitschri/t der Osterrehchischen Cesellseha/t fur Meteoroto^'e, Ti^nA 

 XIV., 1879, s. 250, '• Ueberdie Bestimmung der wahren Regensmenee mi^te'st 

 hochanfge<;telUes Regen'ime^ser, von F. F. Nipher : Meteorologische Zcit- 

 schri/t, Band i., 1884, «• 38', " Uebcr den von Nipher vorgeschlag'nen 

 Srhutztrichter fiir Regenmes<er," v^n R. Bornst°in ; Bulletin de l' Academic 

 Imfihiale des Seiences de St. Petersbour^, v^ s^rie. vol. iii.. 1895. p. I9?i 

 "KritischeUntersuchungder Angabenfreier und geschiitzter Regenmesscr,' 

 von Emil Berg. 



