352 



NA TURE 



[August 8, 1901 



first, to discern these interesting markings. The largest spot 

 seems to be Jrifting westwards, relatively to the zero meridian 

 (System II.) of Mr. Crommelin's ephemerides, at the same rate 

 as the S. temperate zone of Jupiter, viz. i5 degrees per month. 

 On August 3 I observed the spot in transit at Sh. 56m., hence 

 its longitude was 232°-S. If the westerly drift of the object is 

 continued and its existence sufficiently prolonged, it will arrive 

 at conjunction with the following end of the red spot in July, 

 1902, the longitude of the latter being about 46°. This value 

 has shown little variation during the last twelve months. 



The new marking deserves special notice from the fact that a 

 conspicuous spot appeared in about the same latitude in the 

 summer of 1889 and formed the subject of an interesting paper 

 by Mr. A. S. Williams in the Monthly Notices for June, 1S90. 

 This spot, when its more rapid motion enabled it to overtake 

 the red spot, w-as diverted southwards and its material flowed 

 along the southern edge of the ellipse, afterwards drifting further 

 west, freeing itself from the red spot, and finally appearing as a 

 well-defined red stre.ak or short belt. The present disturbance 

 may indicate a recurrence of the phenomena of twelve years ago, 

 and it is desirable that its future developments be closely watched. 

 The chief spot will be central at about the following times :— 



August 10 



Bristol, August 4. 



9 35 



II II 



8 38 

 10 IS 



7 42 



9 18 



W. F. Denning. 



MEASUREMENTS OF SOLAR RADIATION.^ 



IT is impossible within the limits of a short notice to 

 give any detailed description of Prof. S. P. Langley's 

 great work, contained in his first volume of the "Annals 

 ■of the .^sti'ophysical Laboratory of the Smithsonian 

 [nstitution." It is an account of his experiments with 

 the bolometer, begun on Mount Whitney in iSSi and 

 continued for eighteen years, first at the Allegheny 

 Observatory and, since 1890, at the Smithsonian Institu- 

 tion at Washington. The volume before us deals with 

 this last period ; the results of the earlier period are con- 

 tained in the Report of the Mount Whitney E.xpedition, 

 published as one of the professional papers of the U.S. 

 Signal Service Department 'and elsewhere. Striking as 

 these were, their interest and miportance is entirely 

 eclipsed by the new work. 



Those who were fortunate enough to be at the Oxford 

 meeting of the British Association in 1894 heard from 

 Prof. Langley himself an account of his researches up to 

 that date. The details on which that account was based 

 are given in the new volume, together with the results of 

 a completely new set of measurements, made in 

 1897-98. 



Probably only the few who have used the bolometer 

 can fully appreciate its difficulties. The principle of the 

 research is simple. A spectrum is formed by suitable 

 means and allowed to fall on the bolometer, a very narrow- 

 strip of blackened platinum placed parallel to the lines of 

 the spectrum ; this strip forms one of the arms of a Wheat- 

 stone's bridge, the other arms being a similar strip and 

 two equal bridge coils. A screen is interposed between 

 the spectrum and the bolometer, and the bridge adjusted 

 until the galvanometer is undeflected. On removing the 

 screen the bolometer strip is heated by that part of the 

 spectrum which falls on it ; its resistance is changed 

 and, assuming the battery current to remain constant, 

 the deflection of the galvanometer spot is a measure of 

 the change of resistance of the strip and hence of the 

 energy which it absorbs from the spectrum. 



The spectrum is then made to pass slowly across the 

 strip ; by noting the corresponding deflections of the 



' "Annals of the Aslrophysical Observatory of the Smithsonian Institu- 

 tion." Vol. i. By S. P. Langley, Director. 



NO. 1658, VOL. 64] 



galvanometer, a map can be drawn showing the distri- 

 bution of energy in the spectrum. In a region of a great 

 energy the deflection is large, but as an absorption band 

 Crosse's the strip it falls towards zero. When it is stated 

 that a variation in the temperature of the strip of less 

 than one millionth of a degree affects the galvanometer 

 appreciably, some of the difficulties become apparent ; 

 the very slightest changes in temperature of any part of 

 the electrical system produce effects in the galvanometer 

 which mask the effects it is sought to observe. 



In early days the observations were made by eye. The 

 circle carrying the prism or grating which formed the 

 spectrum was set and its position noted ; then the screen 

 was removed and the galvanometer deflection read. 

 The circle was then moved and the operations repeated. 

 In 1891-92 the apparatus was made autographic ; the 

 circle was turned uniformly by means of a clock, thus the 

 spectrum crossed the bolometer strip at a uniform rate, 

 while the spot of light from the bolometer mirror was 

 focussed on to a sheet of photographic paper, which was 

 made to move vertically at a rate bearing a fixed ratio to 

 that of the circle. On developing the sheet a holograph 

 was formed, a curve in which the coordinate of any point 

 measured parallel to the direction of motion of the sheet 

 determines the portions of the spectrum to which the 

 point corresponds, while the coordinate measured at 

 right angles to the direction of motion gives the energy 

 received by the strip from that portion of the spec- 

 trum. Peaks on the curve correspond to maxima of 

 radiation, depressions to absorption bands or minima of 

 radiation. 



A very large number of these bolographs were taken, 

 and the work contains numerous examples. The same 

 general features recur in them, though of course there 

 are small variations from day to day, some of which are 

 real, depending on the weather, while others are acci- 

 dental. A very full account is given of the steps taken 

 to determine which of the depressions were real, i.e. 

 actually due to absorption lines in the solar spectrum, 

 and which were accidental, and of the method of measur- 

 ing the twenty-one bolographs selected for final examina- 

 tion and record. 



Full tables of the wave-lengths of the lines observed 

 and of their relative intensities are given, and the whole 

 is summed up in the normal map of the solar spectrum, 

 which will be found at p. 200. 



The general effect is shown in Plate .\x., which gives 

 the infra-red spectrum of a 60' rock-salt prism from the 

 observations of 1898, showing both the bolographs, or 

 energy curves, and the line spectrum. The bolographs 

 are, it must be remembered, continuous curves having 

 peaks and depressions ; the method adopted for obtain- 

 ing photographically from these a diagram of a line 

 spectrum as usually drawn is explained on p. 73 thus : — 



" One method of obtaining a linear spectrum consists 



in blacking in all of the plate below the photographic 



plate and then photographing this through a combina- 



j tion of a spherical and cylindrical lens, which draws the 



I blacked-in portion out into regions of greater or lesser 



i shade according to the linear depth of the blacked-in 



portion." 



"... In the original curve the greatest elevations 

 represent regions of the greatest heat, the greatest de- 

 pressions regions of the greatest cold ; and if we fix our 

 attention on these great regions only, they can be 

 adequately rendered into bright and dark bands respec- 

 tively, but the detail is comparatively ill rendered with- 

 out a special adjustment, which would in turn give a 

 false presentation of the great masses of light and shade." 



This treatment, illustrated by figures in the book, con- 

 sists of blacking out the whole of the under side of the 

 curve and most of the upper side, but leaving as a trans- 

 parent spot a small area at the vertex of each marked 

 depression, the size and shape of the area depending on 



