June 8, 1882] 



NATURE 



'3i 



oscillation of Nordenskjold's "glory" independence on 

 the seasons. Another, diurnal oscillation, according to 

 which the auroral belt would slowly advance towards the 

 north (for Godthaab) during the night, seems very 

 probable. It would explain — M. Tromholt says- -the 

 greater intensity of aurora towards midnight, as well as 

 the greater frequency of northern auroras among those 

 which were observed at Godthaab during the morning ; 

 but this last phenomenon, of course, might depend also 

 upon some diurnal variation of the intensity of the 

 "common arc." In any case, these conclusions are to 

 be considered as provisory ones, and must be submitted 

 to the further test of observations carried on at points 

 more favourably situated than Godthaab for the study of 

 these oscillations. Such is also the opinion of M. 

 Tromholt himself. 



Of course, the fifteen years' observations at Godthaab 

 do not include a period of time sufficiently long for 

 enabling us to deduce from them the laws of periodicity 

 of auroras. But still they allow of several interesting con- 

 clusions which may serve as a guidance for further re- 

 searches. Thus, it appears from them — contrary to what 

 was said as to the aurora? being more frequent during the 

 most cloudy days — that the number of observed auroras 

 is directly proportionate to the brightness of the sky. 

 This dependence appears not only for different years or 

 months, but also for separate days. If all the days when 

 auroras were observed are classified according to their 

 brightness, which is expressed by the figures 1 to 4, and 

 the brightness compared with the average number of 

 auroras observed during the days thus classified, we see 

 that while the quantity of clouds was 1 6. 17, I 'S, rS ... 

 3"2, 3'3i 3 4, and 3'j. the average corresponding number 

 of auroras was 70, 70, 50, 50 ... 29, 27, 35, and 15, 

 the decrease being altogether very regular, so as to leave 

 little doubt as to the accuracy of the law. 



The following data have some bearing on the 1 \\ years' 

 period of auroras which was deduced from observations 

 in more southern latitudes, and which is considered as 

 depending upon the amount of solar spots. Reckoning the 

 years from August to May, so as to comprise in each year 

 all autumn, winter, and spring auroras (during the bright 

 nights of the summer they are not observable), the yearly 

 number of auroras during the years 1S65-66 to 1S79-80 is 

 given by M. Tromholt as follows :— 97, 112. 65,84,45, 

 Si, 32. 47, 73, 97. 97, 104, 69, 100, and 75, that is, rather 

 irregular. Nevertheless, it is easy to perceive in these 

 figures a certain periodicity with three maxima corre- 

 sponding to the years 1S66-67, 1S76-77, and 1S7S-79. By 

 introducing a correction which depends upon the bright- 

 ness of the sky, and reducing the observed number of 

 auroras to an average cloudiness, M. Tromholt finds 

 another series which is more in accordance with the 

 number of solar spots as given by Wolf. Both series 

 for the years 1865-66 to 1879-S0 (August to May), appear 

 as follows : — 



Number of aurora, with correction for brightness of sky 



862, 91-3, 67-4, So-9, 517, 56-5, 32 o, 46-0, 78-4, 97-0, 95-0, 



1020, 73 o, S5'2, S3 3 



Number of so'.ar spots 



23-5, 61, 183, 6o'i, 1070, 1335, 9S - 6, $9'4, 517, 32-1, 11 -6, 



13-5, 6 8, 22, 163 



It would seem from these two series, that instead of 

 being proportionate to the number of solar spots, the 

 number of auroras is rather inversely proportionate to 

 this number, the two maxima of auroras corresponding 

 with the two minima of solar spots, and the minimum of 

 auroras arriving one year later than the maximum of solar 

 spots. The same appears still better from the observa- 

 tions at Stykkisholm in Iceland, which run through the 

 years 1846-47 to 1872-73. Both curves for this place 

 (auroras and solar spots), although showing several irre- 

 gularity's, nevertheless display a marked connection 



between the two phenomena ; both inflexions of the 

 auroras curve towards a maximum correspond very well 

 with the minimi of solar spots, and vice versa. The 

 result for Godthaab and Stykkisholm is thus the inverse 

 of what was found in more southern latitudes ; and, to 

 explain this contradiction, the author admits that the 

 " auroral belt" is subject in its oscillations to a period of 

 about eleven years, during which it advances more 

 towards the north at the time when the number of solar 

 spots reaches a minimum, and returns back towards the 

 south during the maximum period of solar spots. 



As to the number of auroras observed respectively 

 during the evening and during the morning, the observa- 

 tions at Godthaab fully confirm the fact already noticed 

 at other places, namely, that auroras are more frequent 

 during evening hours. But it still remains to investigate 

 in how far this difference depends upon the hours'of ob- 

 servation, the observer usually taking notice of nearly all 

 auroras which appear before midnight, and not noticing 

 those which appear during the first six hours after mid- 

 night. 



Such are the questions discussed in M. Tromholt's 

 memoir. As will be seen, they are rather indicated than 

 definitely solved ; but we must be thankful to the author 

 for having raised them, and express a hope that the ob- 

 servations of auroras which are now made to such an 

 extent in Norway and Greenland, may be extended to 

 the polar parts of Siberia and North America ; we 

 earnestly hope that the Meteorological Commission of 

 the Russian Geographical Society, which already has 

 done so much useful work, will soon extend its network 

 of observations over this new field, which becomes every 

 day more and more important. P. K. 



ILLUSTRA T/OA'S OF NE IV OR RARE A MM A LS 



IN THE ZOOLOGICAL SOCIETY'S LIVING 



COLLECTION^ 



VIII. 

 20. T^HE MULE DEER (Cariacus macroHs).— While 

 ■»■ the Virginian Deer (Cariacus virginianus) is 

 widely distributed all over the continent of North 

 America, it is necessary to go far to the west before we 

 arrive within the limits of the range of the two other 

 species of the same group — the Mule Deer (C. macro- 

 tis), and the Black-tail (C. columbianus). Of these 

 western deer, the latter, of which the Zoological Society 

 had living specimens some years ago,- is confined to a 

 narrow strip of land along the Pacific coast. But the 

 Mule Deer has a larger distribution, being found on both 

 sides of the Rocky Mountains, and extending eastwards 

 of the main range, far into the prairies of Missouri. 



The Mule Desr was discovered by Lewis and Clarke 

 during their expedition to the Rocky Mountains in 1804, 

 on the Missouri River, in about 42 : N.L., and was so 

 named from the excessive development of the ears, which 

 at once distinguishes it from its fellows. Its most natural 

 home is the mountainous region which flanks the main 

 range of Nonh America on both sides, thou 6 h, as already 

 stated, it extends hundreds of miles into the great plains 

 drained by the Mississippi and its affluents. It is also 

 met vi ith in C regon and British Columbia, though rather 

 superseded in numbers in this quarter by the Black-tailed 

 Deer. 



The antlers of the Mule Deer, which, as in most other 

 deer, are borne only b> the male, are of the same peculiar 

 type as those of the Virginian Deer. All the normal 

 tines have a posterior projection, and the beam, after 

 casting off the basal snagg, curves gradually forward and 

 inward, until the extremities remotely approach one 

 another. The tines thus stand mostly upright when the 

 head is carried in its usual position, but when the head 



1 ' C mit-ued frum vol. xxv. p. 610. 

 = See Wolf and ScUter 'Z.ological Sketches." tvl, i. pi. ao, let hgures of 

 the Deer. 



