294 



NA TURE 



\yan. 28, 1886 



unbroken, was copiously coated with ice crystals above, but only 

 scantily underneath ; but where the ice had been broken so that 

 the cavity beneath it had communicated freely with the open 

 air, the crop of ice crystals was equally copious on both sides. 

 A large hollow iron roller, 24 inches in diameter, had a copious 

 crop of ice crystals on the upper surface, but little on the sides. 

 Underneath, however, it had a coating of ice which extended 

 about an inch from the point at which it rested on the earth, 

 and then ended abruptly. 



These observations seem to show that for the formation of 

 hoar-frost, and inferentially also of dew, upon solid bodies, two 

 factors are necessary, viz. (i) a sufficient cooling of the bodies 

 by radiation below the temperature of the surrounding air ; and 

 (2) a supply of watery vapour in the air ; and that within 

 certain limits an excess of one of these factors may compensate 

 for a deficiency of the other. Thus upon the top twigs, on the 

 uppermost stones in the heaps, and on the upper surface of the 

 iron roller — all these bodies being freely exposed to radiation, 

 and not directly connected with the warm earth — a copious 

 deposit of hoar-frost took place ; and this deposit, it is to be 

 remarked, must have been condensed from the vapour diffused 

 in the layers of air in contact with those bodies, and not from 

 that issuing from the earth. Vapour from the ground could not 

 pass through the impervious iron roller to reach the upper 

 surface ; it may indeed be supposed to have passed round the 

 circumference, but the sides of the roller were, as I have said, 

 comparatively free from hoar-frost. On the other hand, the 

 under sides of the stones, and of the roller, though to a consider- 

 able extent protected from radiation, had on them a deposit of 

 ice, seemingly condensed, as Mr. Aitken supposes, from the 

 vapour exhaled from the comparatively warm ground beneath 

 them. 



The scantier deix>sit of hoar-frost upon the under surface of 

 the "cat's ice" when unbroken than when broken is, I 

 presume, due to the more rapid cooling of the ice when the 

 space beneath it communicates with the open air, than when 

 closed. II. F. P. 



January 19 



Clouds and Upper Wind-Currents over the Atlantic 

 Doldrums 



The first of the two following weather sections across the 

 Atlantic doldrums was taken in June and July last on board 

 the s.s. Toiigariro, during her voyage from Rio Janeiro to 

 Teneriffe. 



Practically clouds in these latitudes may be taken as belonging 

 to three levels : a small cumulus, low down ; a middle layer of 

 some stratiform cloud ; and a high-level cirrus. Any one of these 

 may appear by itself, or all may be present simultaneously. 

 The depth of the various air-currents which drive these clouds, 

 I take to be of great importance in any general theory of the 

 circulation of the atmosphere in the equatorial regions. 



Space will not permit me to give 'here the details of each 

 day's observations, but the results may be briefly stated thus : 

 South of the equator, the low or middle clouds over the south- 

 cast trade, which we picked up in 10° S. lat., invariably came 

 from some point to the right of the surface-wind, when you stood 

 with your back to it, i.e. if the surface-wind was south-east the 

 clouds would drive from about east-south-east. This is the 

 usual rotation of upper currents in the southern hemisphere. 



But, north of the Line, when for reasons which cannot be 

 discussed here, the south-east trade did not turn to south-west, 

 as might have been expected, the upper currents continued to 

 follow the rotation of the southern, and not that of the northern 

 hemisphere, that is to say, the upper currents over the south- 

 east surface-wind continued to come from some more easterly 

 point. In the " doldrums," also, which extended from about 

 8° to 13° N., the same rule obtained, and the middle cloud-layer 

 over some "cat's-paws" of south-east wind drove from the 

 east. 



In the north-east trade T only got one unsatisfactory observa- 

 tion in 22' N. , 19° W., which gave a middlelayer of north-north- 

 east wind over an east-north-east surface trade. This is contrary 

 to what might have been expected. 



The second section was taken in December last on board the 

 s.s. Drummoitd Castle, during part of her voyage between Lis- 

 bon and Cape Town, with much better appliances for observing 

 clouds than on my former voyage. 



In the north-east trade, from 30° N. lat. down to the doldrums 

 in 5° N., the upper layers of cloud invariably came from some 



point to the left of the surface-wind. When you stood with 

 your back to it, i.e. if the surface-wind was from north-east, the 

 higher clouds would come from east, or south-east, or even south 

 by west. This is the usual rotation of upper wind-currents in 

 the northern hemisphere. As far north as 20° N. the middle 

 clouds came from south by west, and in 10' N. this current had 

 descended to the level of the low cumulus, and the middle clouds 

 drove from west. 



But as we touched the doldrums in 5° N., a totally different 

 wind-system became apparent. Over the oily calm of that 

 district I could just detect, through the universal haze and gloom, 

 a middle current from the east ; and when, in a few hours, we 

 picked up the south-west monsoon of the Gulf of Guinea, here 

 coming from south by we^t, the low clouds drove from south- 

 east. This continued till we reached the Line, and the single 

 observation which I got of high cirrus in 1° N. lat. showed an 

 easterly current at that level. Thus, for 8° N. of the equator 

 the rotation of the upper winds was that proper to the southern 

 hemisphere, for south-east and east were over a south-west sur- 

 face-wind instead of west or north-west, as might have been 

 expected. This is the more curious, because the surface-wind 

 has the south-west set proper to the northern hemisphere. 



But the greatest interest of this last observation is to be 

 found in the extraordinary analogy which the wind-system over 

 the Gulf of Guinea presents to the wind-system over the north- 

 west monsoon in the Indian Ocean, which I discovered last 

 spring, and described in a letter to Nature of October 29, 1885 

 (p. 624). In that region I found that as the north-east monsoon 

 crosses the Line and turns to the north-west, the upper currents 

 are those proper to the northern hemisphere, that is to say the 

 low and high layers of cloud come from north and east respec- 

 tively. Now we see that both in the Gulf of Guinea and mid- 

 Atlantic, as the south-east trade crosses the Line, it carries the 

 rotation of the higher currents, proper to itself, up to the 

 doldrums. 



After crossing the equator in the Drummond Castle the wind 

 turned to south by east or south-south-east, and as far as 18° S. , 

 beyond which we need not follow them, the upper currents were 

 either identical in direction with the surface-wind, or else a very 

 little more easterly — that is, they followed the normal rotation of 

 their hemisphere. 



In conclusion, I may point out another very important result 

 of these observations as far as they have gone yet : — The highest 

 current between the equator and the doldrums is always from 

 some point near the east, whatever the westerly set of the sur- 

 face-wind may be. I have had no opportunity of estimating 

 the velocity of this current, but the impression on my mind has 

 been that it is not apparently very rapid, though, of course, the 

 height is very considerable. Ralph Abercromby 



Cape Town, December iS, 18S5 



Ventilation 



Having read your late article on ventilation, and letters 

 referring thereto, I think it may be of interest to your readers 

 to know something of what we have done in this quarter. Since 

 the year 1877 we have had in Dundee a num.ber of schools and 

 other buildings ventilated and heated by propelling large volumes 

 of heated air into the rooms with a small-power engine. 



The system, as now generally introduced, is as follows : — 

 Rotaiy air-pumps, or a Blackroan air-propeller, driven by a 

 small gas-engine, are placed in the basement floor. The air is 

 conducted thither by flues from an altitude of 30 feet, so as to 

 be as free from dust and other impurities as possible. This cold 

 air is discharged into large wooden flues under the ground floor, 

 which are painted with asbestos paint. Running parallel to 

 these flues are others containing hot-water high-pressure pipes, 

 about 1 5 inches in diameter. These hot flues are divided into 

 air-tight sections, suitable for the heating of the different rooms. 

 The air from the cold flue passes into the difterent sections of 

 the hot flue, from which again it passes out, heated, into the 

 different rooms, about 5 feet from the floor, by long and narrow 

 flues, so that it may be well diffused. Each room has its own 

 exhaust-flue or flues, reaching within l| feet of the floor, and 

 passing into the space between the upper-floor ceiling and the 

 roof. This space is made into a large flue, into which all the 

 foul air from the difterent rooms passes. On the roof there are 

 constructed one or more square shafts, with fixed louvre boards 

 on each of the four sides. Inside the louvres are fixed valve- 

 frames covered with a large number of light waterproof cloth 



