August 30, 1877] 



NATURE 



369 



folium, Geranium, Scdum, &c. The forests which clothe 

 the banks of the Jenissei consist to a great extent of non- 

 Scandinavian trees, namely, of the larch, Piiius ccmbra, 

 Aliis sibirica, and the Abit-s ohovata, which scarcely 

 difftrs from Pin us abit-s, and of Populus nigra. Of the 

 trees common to Scandinavia the most important are the 

 birch {Bctula vcrruiosa and glutinosa), pine, aspen, bird's 

 cherry, and mountain ash. Besides, the Saliccs play a 

 very important part on the Jenissei, inasmuch as they 

 form extensive woods on the low banks periodically over- 

 flowed ; these Saliirs often grow to uncommonly large 

 sizes, and consist in part of non-Scandinavian species, 

 one of which, Salix I'itcllina, is the most common of all, 

 and spreads over the whole of the region examined by the 

 botanists of the expedition. 



The bush vegetation too in Siberia exhibits differences 

 from that of Scandinavia. On the Jenissei Alnastcr fru- 

 ticosus is important beyond others, because it forms 

 thickets, and especially towards the north increases in 

 mass, going in that direction beyond the limit of trees. 

 Among other bushes foreign to our flora there occur on 

 the Jenissei the Siberian pea tree {Robinid), Spinta con- 

 fusa, sorbifolia and salicifolia, Sanibuctis raccmosa, Cra- 

 la-gus sanguint\i, Cassandra calyculata, peculiar types of 

 roses, &c. Among the bushes common to Scandinavia 

 the most important are the black and red currant, dwarf 

 birch, Loniicra carulia, which is far more widely distri- 

 buted than in Sweden, juniper, J/i'r/Vi? gale, raspberry, 

 Empctrunt nigrum, Vaccinium vitis ida-a, and myrlillus, 

 &c. Towards the north the bushy Saliccs play an im- 

 portant part, as in our northern regions. On the Jenissei 

 there has been found only one species oi Alnus, which is 

 specially interesting as not being either of the Swedish 

 species, but perhaps the genuine Alnus pubcsccus. On 

 the other hand there are absent on the Jenissei many of 

 our trees and bushes, as the nobler deciduous trees and 

 fruit trees, and, what may be said to be distinctive of the 

 Jenissei flora, heather, which is so extensively distributed 

 in Sweden, is wanting. 



TEMPERATURE AND HUMIDITY OF THE 

 AIR A T DIFFERENT HEIGHTS 



A MEMOIR on the temperature and humidity 

 of the air at different hours, by Dr. H. E. 

 Hamberg, based on observations made by him during 

 the summer of 1S75, at heights varying from 2 inches 

 to 22 feet above the ground, was published recently 

 in the Transactions of the Royal Society of Sciences 

 at Upsal. The memoir is a valuable one, and is of in- 

 terest to more than the mere meteorologist, it being evi- 

 dent that the inquiry is so handled as to bring it into close 

 connection with such difficult questions as convection 

 currents in the free atmosphere and the diffusion of 

 vapour through the air. 



In clear weather the temperature of the air nearest the 

 surface was lower than that above it, from two to three 

 hours before sunset to at least two to three hours after 

 sunrise. At all the six heights the temperature fell to the 

 minimum at the same hour, viz., about 3 a.m. ; but while 

 it continued from this time to rise steadily at all the 

 heights, the lowest temperatures continued to be observed 

 in the strata nearest the ground till several hours after 

 sunrise. From this remarkable result Dr. Hamberg con- 

 cludes that the increase of temperature in the lower strata 

 of the air in the early part of the forenoon is not an im- 

 mediate and direct consequence of the heating of the 

 ground, but is rather to be attributed to the absorption by 

 the air, or more strictly by its aqueous vapour, of the heat 

 received from the sun's rays or reflected from the ground. 



Over uneven ground covered with vegetation the tem- 

 perature near the surface is generally higher over those 

 parts of the field which rise above the general level. 

 Thus even slight elevations of only one or two feet have 



the air immediately resting on them often 2° higher or 

 more, whilst on the other hand, a trench or depression 

 one or two feet below the general level has the air rest- 

 ing on it often 2°, or more, lower than the air over the 

 level portions of the field, a result of considerable 

 practical importance in agriculture and horticulture. 



The latent heat set free on the formation of dew appears 

 from the observations clearly to retard the lowering of 

 the temperature, but not to the extent which might have 

 been expected. When, on the deposition of dew, the 

 temperature of the air near the surface has fallen below 

 32°, as soon as the dew is congealed into hoar-frost the 

 temperature of the lowest stratum of air in contact with 

 the ground instantly rises to 32° ; but at the same time 

 the temperature of the air higher up steadily remains 

 lower than 32°. 



The absolute humidity of the air on clear nights on 

 which no dew is deposited decreases from the ground 

 upwards, just as happens during the day ; but on the 

 other hand, with dev/, the humidity is least nearest the 

 ground, and increases with the height, and this influence 

 of dew, in diminishing the humidity, extends [upwards to 

 at least twenty-two feet, the height to which the observa- 

 tions were carried. Since his observations clearly show 

 that the absolute humidity begins in the evening to 

 diminish near the ground before any dew is observed to 

 be deposited, and also diminishes at all heights on those 

 nights during which no dew whatever is formed. Dr. 

 Hamberg is of opinion that the diminution of the humidity 

 of the air during night is to be sought for in other 

 physical causes than the deposition of dew. 



Several of the points discussed will doubtless be made 

 subjects of further investigation by others. In all cases 

 it is most desirable, indeed absolutely necessary, to a 

 critical valuation of the observations, that the authors 

 give woodcuts and descriptions of the exact position and 

 mode of protection adopted in the case of each thermo- 

 meter employed in the observations. For such refined 

 inquiries the method of observation must necessarily be a 

 refined one ; in other words, such as will certainly secure 

 the necessary comparability among all the instruments. 



THE CHRONOMETERS OF SWITZERLAND 

 VXTE find in a recent number of the Bulletin of the 

 • "^ Society of Natural Sciences at Neuchatel an 

 interesting report of the Neuchatel Observatory on the 

 annual competition of chronometers for prizes awarded 

 yearly by the Observatory. The report gives some idea 

 of the degree of perfection reached in Switzerland in the 

 construction of chronometers. The 231 chronometers 

 (four box and 237 pocket chronometers) admitted to the 

 competition out of 270 sent in were submitted to a severe 

 test, including changes of temperature and of position 

 during periods of from two weeks to two months, and the 

 results of the trial appear as follows. The average 

 diurnal variations in the rates of the box-chronometers 

 proved o'32 sec. (against o'20 in 1874), and of the pocket 

 ones 0*46 sec. (against o'53 in 1874), and there was but 

 two per cent, of these last, the average diurnal variation 

 of which reached i sec. The various systems of es- 

 capements appeared, as was observed in former years, 

 to have but little if any influence on the degree of pre- 

 cision of watches, provided they remain constantly in the 

 same position, vertical or horizontal. The compensation for 

 changes of temperature proved to bealtogether satisfactory, 

 the average variation of 167 chronometers submitted to 

 variations of temperature from 15 to 25 Cent, degrees being 

 but o'i3 sec. for each Centigrade degree. It must, how- 

 ever, be noticed that only 51 per cent, of them showed 

 variations below one- tenth of a second for each degree, 

 and that 10 per cent, showed variations above o'3 sec. 

 Finally, the differences between the rates during the first 

 and the last weeks were : for bo.x- chronometers, 2'il sec. 



