55^ 



NATURE 



{Oct. 19, 1876 



was brought prominently to light during the time that the 

 normal sidereal clock for Greenwich was under trial. 

 That clock had been fitted with a heavy mercurial pen- 

 dulum, and it was found that the rod got warmer or 

 colder some time in advance of the mercury ; of course 

 the compensation failed for such interval of time. 



The following form of pendulum was afterwards sub- 

 stituted. The expansion of zinc is, as you see, nearly 

 double that of brass, and consequently a good deal 

 less of it is required to compensate a pendulum. To 

 the extremity of an internal steel rod (see Fig. 12) a 

 collar is fastened, and a zinc tube inclosing the steel rod 

 rests upon it. To the summit of the zinc tube is 

 attached a steel tube, which in turn incloses both it and 

 the rod, and the pendulum bob is fastened midway of its 

 length to the extremity of this tube. The outer steel tube 

 is cut away at its sides, and holes bored in the zinc in 

 order to let in changes of temperature rapidly. 



The action of the combination is similar to that of the 

 gridiron pendulum, the expansion of the zinc upwards 

 exactly neutralising and destroying the expansion of the 

 steel downwards. It is important (as suggested by Mr. 

 Buckney) that the bob should be suspended at its centre, 

 because otherwise it would also operate as an expansion 

 length, and although its effect could be counterbalanced 

 by shortening the zinc tube, yet owing to its bulk it would be 

 sure to lag behind the rest of the compensation, and cause 

 such an error as I have referred to. 



Barometric Compensation, 



When you aim at the very highest time-keeping, baro- 

 metric compensation becomes necessary ; that is to say, 

 compensation against the disturbance to the pendulum 

 due to changes of atmospheric pressure. For instance, 

 when there is any rise in pressure, when the atmosphere 

 becomes denser, our clock will lose, and will gain when 

 the atmosphere becomes moi« attenuated, the variation 

 in the Greenwich clock having been at about the rate of 

 •3 of a second a-day for a difference of one inch in the 

 bare meter. 



The following compensation (see Fig. 13) is one de- 

 signed by Sir George Airy : — 



C is a lever moving around an axis at A. One arm of 

 the lever carries a horse-shoe magnet, b, and the other a 

 float, e, supported upon the mercury in a barometer cistern. 

 Two bar magnets (the front one, a, only is shown, the 

 other being behind the bob) are fastened upon the pen- 

 dulum bob, the north pole of one pointing upwards, and 

 of the other downwards (in order to render the combina- 

 tion astatic). 



The poles of the horse-shoe magnet face the opposite 

 Tioles of the two bar magnets, and attraction goes on 

 ictween them. When the barometer rises the mercury 

 in the cistern falls, and with it the float. The other arm 

 of the lever, therefore, rises, bringing the poles of the 

 hot so- shoe magnet closer to the poles of the two bar- 

 magnets, and increases the attraction between them, 

 which is a force acting in the same direction as gravity. 

 The pendulum consequently moves faster (for we increase 

 the pull upon it), the tendency to go slow arising from 

 the increased atmospheric pressure is by this means com- 

 pensated. Dr. Robinson, at the Armagh Observatory, 

 effected the same correction by attaching a barometer 

 to the pendulum rod. He also noticed that changes in 

 atmospheric pressure would disturb a mercurial pendulum 

 to a very considerable extent if there were air-bubbles in 

 the mercury. 



{Ts be cfifitinued.) 



CROOKES'S RADIOMETER 

 T HAVE recently made a few experiments with this 

 •*■ instrument which may not be uninteresting to the 

 readers of Nature. 

 The radiometer used had discs of aluminium polished 



on one side and blackened on the other ; it was more 

 than usually sensitive, and would sometimes continue its 

 rotation for twenty minutes after the sun had set in the 

 sea. 



The instrument being in a room in which the radiation 

 was far too feeble to cause the arms to move, I grasped 

 the bulb with both hands, so as still further to exclude it 

 from light. The vanes immediately began to revolve 

 briskly, the polished sides first. Removing my hands 

 after two or three minutes, the movement soon stopped ; 

 and then, after a very brief interval of rest, began in 

 the opposite direction, and so continued for several 

 minutes. 



I now placed the instrument in a room, near to a window 

 through which the light of the full moon in a clear atmo- 

 sphere was shining. The arms of the radiometer did not 

 move. By means of a large lens the moonlight was then 

 concentrated about 200 times, and allowed to fall full 

 upon the blackened side of one of the circular discs, in 

 such a way as to cause the intensely brilliant image of 

 the moon to nearly cover the disc. Not the slightest 

 movement occurred, although the concentrated light im- 

 pinged upon the disc for a quarter of an hour. 



As is well known, the light of the moon contains, for a 

 given luminosity, far less heat rays than does light from 

 any terrestrial source, no matter how much the latter 

 may be strained through intranscalent media ; in fact it 

 require Lord Rosse's 6-feet reflector clearly to demon- 

 strate the excessively feeble thermal power of the lunar 

 rays. 



These experiments show, firstly, that light is not 

 necessary to the movement of the radiometer ; secondly, 

 that light only contributes to the movement in so far 

 as, by its absorption, it is transformed into heat ; ,and 

 thirdly, that the motion is due to the unequal heating 

 of the two sides of the discs, the cooler surfaces always 

 preceding the warmer ; for when the instrument was 

 grasped by the hands, the blackened surfaces of the 

 discs rapidly absorbed the heat rays, whilst the polished 

 surfaces reflected them. Thus the surfaces of the 

 blackened discs remained warmer than the metal beneath, 

 but gradually communicated their heat to the latter. On 

 removing the hands from the bulb, the thermal condition 

 of the discs would soon become reversed ; the black 

 surface — a good absorber and also a good radiator — 

 would cool much faster than the opposite surface, A^hich 

 being of polished metal was an exceedingly bad radiator. 

 The blackened surfaces, therefore, now became the 

 coolest, and preceded the polished ones, in other words, 

 the direction of rotation became reversed. 



October 17 E. Frankland 



THE GEOLOGY OF ENGLAND AND WALES^ 



THE well-known volume of Conybeare and Phillips, 

 entitled " Outlines of the Geology of England and 

 Wales," which was published in 1822, and was based on 

 an earlier and slighter work of the second-named author, 

 has long held an honourable place among geological 

 classics. It has served, indeed, to supply to some extent 

 the want so universally felt of a descriptive memoir or 

 handbook to William Smith's Geological Map, a work 

 which " the father of English geology " could never be 

 prevailed upon to write himself. The " Outlines," how- 

 ever, is but a fragment, the second part of the work, 

 which was to have dealt with the oldest rocks and with 

 questions of Economic Geology, never having been pub- 

 lished ; and more than half-a-century of research, carried 

 on in connection with a science which appears to have as 



I " The Geology of England and Wales : a Conci-e Account of the Litho- 

 logical Characteis, Leading Fossils, and Economic Products of the Rocks ; 

 with Notes on the Physical Features of the Country. By Horace B. Wood- 

 ward, F.G.S., of the Geological Survey of England and Wales. (London : 

 Longmans, Green, and Co. , 1876 ) 



