NATURE 



265 



THURSDAY, JULY 20, 1905. 



ARCTIC METEOROLOGICAL OBSERVATIONS. 

 The Norwegian North- Polar Expedition, 1S93-1896; 



Scientific Results. Edited by Fridtjof Nansen. 



Vol. vi. Published by the Fridtjof Nansen Fund 



for the Advancement of Science. Pp. xiv + 659; 



20 plates. (London : Longmans, Green and Co., 



1905.) Price 365. net. 

 T T is a misfortune that meteorolog-ical results de- 

 mand so many figures for their discussion, and 

 so much space for their exhibition. Vol. vi. of the 

 Norwegian North Polar Expedition, dealing with 

 the climatology of the area through which the Frain 

 drifted in its memorable voyage, 1893-6, is a most 

 interesting book, but its size and appearance might 

 repel any but the most ardent meteorologist. The 

 discussion of the observations has been undertaken 

 by Prof. Mohn, of Christiania, and the arrangement 

 is a model of clearness and efficiency. Prof. Mohn 

 superintended the whole meteorological equipment, 

 suggested the plan of work to be carried out during 

 the voyage, and arranged with Captain Scott-Hansen 

 the general management of the meteorolog'ical work. 

 We imagine Prof. Mohn must be gratified with the 

 success of his arrangements, and the intelligent 

 interest which the officers of the expedition have 

 shown in the work. Notwithstanding the severity of 

 the climate, there is hardly a gap in the series of 

 observations. At sea, the observations were taken at 

 intervals of four hours, but for the greater part of 

 the time the readings were made every two hours, 

 with a regularity that compels admiration. The 

 result is that we have, with very considerable accuracy, 

 the climatological elements of a region in the circum- 

 polar Arctic Ocean, where the surface of the earth 

 during the whole time was of a unique homogeneous 

 nature, consisting of a level of frozen water, remote 

 from continents and islands, and with an uninter- 

 rupted free horizon. 



The wind, particularly with regard to direction and 

 velocity, is the first element discussed. To obtain a 

 sufficiently long series of observations for investi- 

 gation. Prof. Mohn divides the interval into three 

 groups, a dark season when the sun was below the 

 horizon, a sunny season during which the sun was 

 above the horizon for practically twenty-four hours, 

 and the equinoctial months, during which there was 

 regular day and night. The discussion shows that 

 during the dark season the wind shifts generally 

 against the sun. Only during four hours in the 

 twenty-four does the wind veer with the sun, while in 

 the sunny period the wind veers with the sun, backing 

 about six hours, divided into periods of two hours 

 each at three different periods of the day. In the 

 equinoctial months the backing and veering are 

 equal, the wind shifting with the sun during the night 

 and morning, and against the sun from 10 a.m. to 

 10 p.m. The diurnal period of the wind's direction 

 is a phenomenon which still awaits an explanation, 

 and the different direction of the shift of the wind in 

 NO. 1864, VOL. 72] 



the dark and in the sunny season seems to be of some 

 importance for the solution of the problem. With 

 reference to the velocity of the wind, it is shown to 

 be greater when the sky is overcast than when it is 

 clear. In the former case, the average velocity is 

 5.09 metres per second (11-4 miles per hour), and in 

 clear weather only 3-54 metres per second (8 miles per 

 hour). The greatest velocity recorded appears to be 

 40 miles an hour in February, 1S96. 



The discussions of the variations of temperature are 

 very interesting, but the results drawn from them re- 

 garding the periods of the meteorological elements 

 must of necessity be less trustworthy than if there had 

 been a longer series of observations at disposal. It 

 may therefore be premature to draw conclusions as to 

 the connection between the different observed pheno- 

 mena, and between those phenomena and their prob- 

 able causes. The desirability of a longer period, and 

 the character of the errors that can be introduced by 

 the comparison of but few values, are shown very 

 readily if we attempt to derive the month of lowest 

 temperature from the figures given. The readings 

 are centigrade, and show the mean temperature for 

 each month : — 



February March April 



J^ 



1895 



■35-83 



7S 



The great variation of temperature in March, 1896, 

 making it nearly equal to that of April, demonstrates 

 the uncertainty that must accompany any attempt to 

 derive mean values from short periods. But the de- 

 ductions drawn directly from the observations, and 

 supported as they are in many instances by similar 

 observations made in Arctic latitudes, are not liable 

 to the same uncertainty. Among these results may 

 be placed the following : — Throughout the dark winter 

 months, when the sky is clear, the lowest temperature 

 occurs in the day, the highest during the night. 

 Generally, in the other months, we have the ordinary 

 diurnal period. With the sky overcast, the diurnal 

 period, with a minimum in the early morning hours 

 and ma.ximum after noon, is very well developed in 

 all the months except January. 



"The most striking feature," says Prof. Mohn, 

 " seems to me to be the distinct diurnal period of the 

 ordinary march in the winter and dark season, with 

 the sky overcast and relatively higher temperatures. 

 The inverted period with clear sky in the dark season 

 seems to be due to the diurnal period of the wind's 

 direction. The dark-season period with its stronger, 

 south-easterly winds, is hardly to be accounted for by 

 the radiation from the sun or sky." 



The forms of cloud, the relative humidity, and the 

 amount and character of precipitation are discussed at 

 full length, but do not present results of unusual 

 importance. With regard to the latter, however, it 

 is not altogether uninteresting to notice that the 

 number of days in a year on whic-fi rain is probable 

 is 49, while snow may be expected on 157 days, and 

 some form of moisture will be collected on x8o days. 



N 



