April 21, 1887] 



NA TURE 



589 



rainband is that its intensity is the same in every possible 

 direction, whereas in clear weather, as is well known, it 

 is generally greatest at the horizon and least at the zenith. 

 But as the rainband in fog or mist has been found to be 

 equal to that which is observed in clear weather at an 

 altitude of 30° or 40° — the altitude of that part of sky 

 always observed — the presence or absence of fog and 

 mist has been ignored in working up the observations. 



In forecasting the weather for the surrounding low- 

 levels, the rainband observed here, together with similar 

 observations at Fort William, would probably be of great 

 value, but its forecasting power for the summit alone is 

 limited. The lower values generally show indications of 

 rain several hours before it comes on, but the higher 

 values simply indicate a continuation of the heavy rain- 

 fall by which they are invariably accompanied. To show 

 that the rainfall increases with the successive numbers of 

 the scale, the mean hourly rainfall has been computed, 

 for each scale value, from the rainfall of the three hours 

 and twelve hours after noon. For the observations of 

 1 8S5, the mean daily rainband was used. In 1 886, each 

 individual observation was taken into account, and, for 

 each number of the scale, the mean rainfall was com- 

 puted from the rainfall of the three hours after each 

 observation. As the results differ in some respects, they 

 are here given for each year separately : — 



Rainband o 12345 ^"^d upwards 



Rainfall f I S85... -000 '006 •016 '029 '050 106 



(inches) \ 1S86...002 '006 ■012 '016 "027 '076 



Owing to the fewness of observations of the numbers 

 5, 6, and 7, they have been grouped together. The higher 

 values are followed by a rainfall which is proportionally 

 far too high, owing, no doubt, to the fact that these higher 

 values are only observed during the passage of cyclonic 

 disturbances laden with moisture from the Atlantic, when 

 a great amount of this moisture only comes into the 

 spectroscopic field in a condensed state, when it is forced 

 to ascend so as to pass the summit, and consequently 

 does not affect the rainband, but causes a very heavy 

 rainfall. The moisture that ascends the mountain, not 

 being detected below by the spectroscope here, is a con- 

 stant source of disparities in the agreement of rainband 

 and subsequent rainfall. The fact that our mean rain- 

 band is not greater than 17, and that the amounts of 

 rainfall were, for 1885, i46'497 inches, and for 18S6, 

 I07'847 inches, clearly indicates that a great part of our 

 rainfall is due to the condensation of the moisture that is 

 forced up from below the level of the summit. 



In comparing rainband with subsequent rainfall, the 

 temperature of the air at the time of observation, as well 

 as the variation in the temperature, must be taken into 

 account. With the view of ascertaining the relations 

 between rainband, subsequent rainfall, and temperature, 

 the mean hourly rainfall for the three hours after an ob- 

 servation, for each number of the rainband scale, and for 

 every 5' of temperature from 15' to 50°, has been calcu- 

 lated, and the results for rainbands of values I, 2, and 3, 

 are as follows : — 



Temperature 15° so° 25° 30^* 35° 40° 45^ 50° 



j^ I ... '003 "O05 010 •C06 006 '004 'OOO 'OOO 



Rainband-] 2 ... 005 'oio 024 '015 012 '006 "003 'ooo 



1,3 ... 007 013 026 -oig 016 015 009 003 



The means for temperatures of 25' and upwards show 

 that for any one rainband, when the temperature rises 

 the rainfall decreases, and when the temperature falls the 

 rainfall increases. The results being as yet only ten- 

 tative, it cannot be definitely stated by how much the 

 mean hourly rainfall increases or decreases per degree of 

 fall or rise in the temperature, for any one value of the 

 rainband. The means for temperatures below 25' seem 

 to indicate that a fall in the temperature causes a decrease 

 in the rainfall, which is not at all probable. If it be the 



case that low temperatures do not affect the absorbing 

 powers of aqueous vapour, which is not likely, the smal 

 amount of rainfall at these low temperatures may be due 

 to the necessarily unsatisfactory measurements of pre- 

 cipitation obtained from the rain-gauge when the temper- 

 ature is below 32°. In truth, snowfall and rainfall as 

 measured by the rain-gauge, can hardly be compared with 

 each other. For a full account of these and other ques- 

 tions of rainfall here, see Mr. Omond's articles and also 

 Mr. Buchan's " Meteorology of Ben Nevis " in the Journal 

 of the Scottish Meteorological Society for the last two 

 years. 



The mean rainbands at the different temperatures 

 are : — 



Temperature 15° 20° 25° 30'' 35° 40° 450 50° 



Rainband ... 04 fi I'l 1-5 19 2-4 2-5 2-3 



The greatest number of observations at any one tem- 

 perature was 724 at 30', and the least 35 at 50°. 



The reason why so many observations were made in 

 1886 was to find whether there was any daily variation in 

 the rainband. That there is a slight variation will be seen 

 from the following results, which are for the summer 

 months only, viz. : — 



Hour 6 7 S g lo ii 12 noon 



R.ainband 1'4 I '4 I '6 I '5 I '6 16 17 



Temperature 32'-o 32'-3 32^7 33'-2 33"-5 34"-o 34°-4 



Hour 13 14 J5 i5 17 18 



Rainband 17 17 I'S 17 I'S 17 



Temperature 34"-S 35"-o 35 = -o 34-°9 34°7 34°-3 



The rise in the rainband from 6 a.m. till 3 p.m. follow- 

 ing the temperature, points to the cause of this daily 

 variation as being the e.\pansion of the lower atmosphere 

 by the rising temperature, and the consequent raising of 

 the vapour above the level of the Ben. This is almost 

 the same cause to which Mr. Buchan ascribes the rise in 

 the barometric pressure for the same daily period. Prof. 

 Piazzi Smyth says that the rainband does not increase 

 for a rise of temperature at sea-lc'cl, because the total 

 quantity of vapour over the place of observation is pretty 

 constant (Journal Scot. Met. Soc, vol. v.). But over this 

 summit the quantity of vapour is not constant, but varies, 

 and the rainband varies with it. 



Remarkable variations in the rainband occur in the 

 course of a single day, often amounting to 3, and some- 

 times to 5 and 6. On March 26, 1886, when a cyclone 

 was passing to the north of the Observatory, the rainband 

 varied from 6 at 10 a.m., to 3 at 12 noon, and to i at 

 5 p.m. The strongest mean for any one day was 8 on 

 January I, 1886, while a mean of o has frequently 

 occurred, en December 18, 1885, a rainband, estimated 

 at 12, was observed on the rising sun — this was one of 

 the noted " fore-glow " mornings. In anticy clonic weather, 

 which is characterised by great dryness of the air, with 

 all the clouds at lower levels, a strong rainband is always 

 obtained from the layer of air close over the clouds, in 

 summer, when not a trace is detected at the usual height 

 of observation ; but in similar weather in winter the rain- 

 band is often entirely absent, even over the clouds. It is 

 noteworthy that on days with little or no rainband, when 

 not actually looking at the sun, the spectrum as a whole 

 is darker than usual, and on days with strong rain- 

 bands, the parts of the spectrum not occupied by lines 

 are brighter than usual. In a certain type of weather, when 

 the rain-gauge completely fails to record the precipitation, 

 the rainband always indicates the presence in the air of 

 the vapour which gives it. This occurs when snow- 

 crystals are deposited with low temperatures and strong 

 winds, as described in N.\TURE (vol. -xxxi. p. 532). The 

 importance of rainband observations will be greatly in- 

 creased when similar series can be undertaken at the sea- 

 level simultaneously with those on the summit of Ben 

 Nevis. A. Rankin 



