714 



NATURE 



[February 25, 1915 



between theory (on the hypothesis of a solid earth) 

 and observation, as regards precession and nutation. 



Newcomb appears to have suggested that viscosity 

 might possibly render precession and nutation the 

 same as if the earth were rigid throughout. As a 

 direct cause viscosity is inadmissible, but indirectly 

 it is effective, for it at once occurred to Lord Kelvin 

 that a very real cause of agreement between an intern- 

 ally liquid earth and a solid earth as regards pre- 

 cession at least, was probably to be found in the 

 rigidity induced in the interior liquid by its rotation. 

 Thus his attention was directed to the quasi- 

 rigidity of a liquid induced by rotational (or vortex) 

 motion, a subject which, as he told Section A of the 

 British Association after his return from America, 

 occupied his thoughts for weeks almost to the 

 exclusion of all other scientific subjects. 



He soon found that if the ellipticity is not too 

 small the shell would not have more precession than 

 the liquid, and that the compound rotating mass 

 would have sensibly the same precessional motion as 

 if it were a single rigid body. He came to the 

 conclusion, however, that the lunar semi-annual and 

 lunar fortnightly nutations would be greatly affected 

 by interior liquidity of the earth. 



At the Glasgow meeting of the British Association 

 in 1876 Lord Kelvin was president of Section A, and 

 began his presidential address by quoting the Anacre- 

 ontic couplet : — 



" 0eX<a \tyiiv ArpetSas, 

 Ot'Aci) Se KaS/xov adfiv" 



which begins the complaint of the poet that no 

 matter what hero he wished to sing, his lyre refused 

 to respond to any theme but that of love. Try as Lord 

 Kelvin liked to speak of the scientific men, and 

 scientific inventions that he saw in America, of 

 American education, or the more recent advances of 

 phvsical science, his thoughts ever came back to the 

 subject of the internal rigidity of the earth and the 

 ■difficult questions therewith connected. So to this 

 topic he decided to devote the major part of his 

 address. This he did with great effect, clearing away 

 what was doubtful from his former arguments, em- 

 phasising and enforcing them as they remained, and 

 reiterating with undiminished confidence his old con- 

 clusions. 



To illustrate the precession of a rotating liquid he 

 showed later in Section A what he called a liquid 

 gyrostat, and also for comparison various solid gyro- 

 stats which had for several years been used for the 

 •dynamical illustrations of the natural philosophy class. 

 I have these very gyrostats here on the table, and 

 will use them for a repetition of some of the old 

 historical experiments of the Glasgow class-room. 



IV. — Solid and Liquid Gyrostats. Gyrostatic 

 Experiments. 



The construction of a solid gyrostat is shown in the 

 •diagrams before you, which were made, partly by 

 myself, nearly forty years ago (Fig. 5). The instru- 

 ment consists of a massive flywheel surrounded by a 

 case of brass. The wheel is a disc of thick brass 

 carrying a massive rim, so that the moment of inertia 

 is made as great as possible. One diagram of this 

 slide, as you see, shows a section of the wheel and 

 case, the other a side view of the wheel. 



The case is a cylindrical box surrounding the f!)'- 

 wheel, with extensions enclosing the axle, for which 

 they are provided with bearings at the ends. Round 

 the case, as nearly as may be in the central plane of 

 the flywheel, is a projecting rim, the edge of which 

 is not quite circular, but rather polygonal with curved 

 sides, and the points of meeting of the sides rounded 



NO. 2365, VOL. 94] 



off. The rim serves to support the gyrostat, as it 

 stands on this glass plate, in some of its evolutions. 



The bearings are cups in which the rounded points 

 of hardened steel of the axle run. This is not a good 

 arrangement if the gyrostat is to be subjected to 

 shocks, or to be roughly handled in any way. Oiling 

 also is required, after every second spin at least. In 

 our new gyrostats we use ball bearings designed to 

 resist considerable shocks and stresses without 

 derangement. With these, in some experiments, we 

 have gone up to speeds of about 25,000 r.p.m., and 

 have found the flywheel to be still rotating rapidly 

 after the lapse of forty-five minutes. Also the wheel 

 may be run for several hours with only one oiling. 



It will be convenient to show here some of the 

 experiments usually performed in the ordinary class 

 of natural philosophy in Lord Kelvin's time. The 

 multiplicity of subjects put down to be treated in 

 the dynamical part of the course precluded, as I have 

 hinted, any detailed explanations of these experi- 

 ments. Thev 



carried 



1 fact, 



t h e 



and 



were 

 out, ii 

 with 

 avowed 



excellent pur- 

 pose of excit- 

 ing curiosity in 

 the minds of 

 the students, 

 and a desire 

 to find out why 

 gyrostats be- 

 ll a v e in a 

 manner at 

 first sight so 

 anomalous. In- 

 terest was cer- 

 tainly aroused 

 in a few, but I 

 fear that the 

 majority 

 despaired o f 

 pe ne t r a ting 

 such mysteries, 

 and sought 

 external help 

 for the master- 

 ing of the more 

 hackneyed 

 topics of the 

 degree exam- 

 inations. 



The process 

 o f spinning 

 excited more interest 

 the experiment, for 

 student cares more 



Fig. s. 



any other part of 

 ordinary elementary 

 little bit of sensa- 



than 

 the 

 for a 



tion than about the scientific result to be proved. A 

 long cord was laid out on the floor, then the free 

 end passed one and a half, or two and a half, times 

 round the axle of the gyrostat, which was held by the 

 operator, with its axis vertical, in a suitable socket 

 on a table fixed to the floor. An attendant holding 

 the free end ran away with it, slowly at first, then 

 faster and faster, down a long passage and through 

 a large adjoining room, while friction was applied 

 to the cord as it entered the gyrostat case. 



For the runner was substituted later a large wheel 

 with grooved rim on which the cord was wound as it 

 was drawn through the gyrostat. I estimate that 

 speeds of about 100 turns per sec. or less may have 

 been obtained in this way. Now, of course, one sp'ns 

 by an electric motor, as I shall presently descr.be. 



