February 25, 19 15] 



NATURE 



711 



level, and the hours of observation. The hours are 

 not given, nor the heights of the places above sea- 

 level, nor the months included in the seasonal periods. 

 All these things may be of no interest to the sailor 

 who will be expected to test the charts by his experi- 

 ence, but they are vital to the utilisation of the author's 

 work by other meteorologists either for research or 

 for incorporation in charts for wider areas. 



The tables of wind give (i) for sixty-six places the 

 "mean direction" of the wind for each season and the 

 year, computed probably from Lambert's or some 

 similar formula ; (2) for seventeen regions of the 

 Mediterranean the frequency of light, moderate, and 

 strong winds for sixteen points of the compass, and 

 the number of calms, also for each season and the 

 5'ear. The regions are roughly about 200 miles square. 



The charts have been drawn with a great deal of 

 care and the delineation is clear and attractive. They 

 are on a relatively large scale, and consequently are 

 folded into the book, which makes it troublesome to 

 consult them, but the reader will feel that his trouble 

 is compensated by the ease of seeing rapidly the 

 general features and the details of the meteorological 

 distribution. It may be remarked, however, that 

 charts of normal distribution intended for the sea- 

 faring man ought to be printed on good linen-backed 

 paper if they are to be used and not folded and put 

 away in a drawer until a new set is received. Indeed, 

 it is doubtful if the ordinary navigating officer will 

 consult meteorological charts regularly until they are 

 made part and parcel of his everyday equipment by 

 being printed on his navigatmg charts. 



Prof. Marini's charts showing the wind roses for the 

 seventeen regions referred to, are very interesting, 

 brinefing out most clearlv the relatively stormy winter 

 conditions of the western Mediterranean and the great 

 preponderance of winds from nearly due east or due 

 west between Sardinia and Gibraltar. No wind rose 

 is given for any part of the Adriatic. E. G. 



LORD KELVIN'S WORK ON GYRO- 

 STATICS.^ 

 I. — Gyrostatic Experiments in the Glasgow Class- 

 room. 

 "\1/'HEN I was a student, and afterwards when I 

 ** was an assistant at Glasgow, Lord Kelvin 

 lectured to his ordinary class twice a week, when he 

 was not called away, and his subject was dynamics. 

 About the middle of the session gyrostats made their 

 appearance on the lecture-table, and we had wonder- 

 ful gyrostatic experiments which filled us with delight, 

 and gyrostatic question^ in the weekly class-examina- 

 tions which were equally productive of dismay. These 

 gyrostatic questions, like many of our exercises in 

 natural philosophy, were often of a numerical char- 

 acter. It is always a good thing to get dow-n to 

 numbers, and it is a most healthful mental discipline 

 to be forced to " get the units right." Our equipment 

 for the solution of these problems was of the slightest, 

 for Lord Kelvin was himself so keenly absorbed in 

 observing the behaviour of the apparatus, that he 

 rather frequently forgot to give us the full dynamical 

 explanation of the curious evolutions which we be- 

 held. I could follow the process of composition of 

 angular momenta, and could see that the axis of 

 resultant angular momentum turned at that rate; but 

 whv should that also be the rate of turning of the axis 

 of figure? That was my special difficulty, and it was 

 only afterwards, when I got the idea of steadv motion, 

 and saw how the general equation is obtained and 

 how it breaks down into the conditions for steady 



1 Abridged from the Sixth Kelvin Lecture, delivered at the Institution of 

 Electrical Engineers, on January «8, by Prof. A. Gray, F.R.S. 



NO. 2^6^, VOL. 94I 



motion, that the matter became clear. Then I found, 

 moreover, that in the general case there are two 

 possible rates of turning. It is a good thing to 

 stimulate the curiosity of a student to make him find 

 out things for himself : it is also an excellent thing 

 to anticipate his difficulties to some extent, lest he 

 grow weary and faint by the toilsome dynamical way. 

 The lectures that we had were undoubtedly most 

 interesting and suggestive, though they were not per- 

 haps always directed to the more prosaic topics which 

 formed the staple matter of the degree examination 

 questions for ordinary students. The first experiment 

 made was always that of the equilibrium of this nearly 

 egg-shaped piece of wood, which, scientifically de- 

 scribed, is a homogeneous prolate ellipsoid of revolu- 

 tion. Its surface may be imagined to be generated 

 by the revolution of an ellipse about its longer axis. 

 • (The diagram. Fig. i, shows a really egg-shaped 

 solid.) I lay it on its side, and we see that in that 

 position it is stable for fore and aft inclinations, 

 "pitching" I may call the motion, and in indifferent 

 equilibrium for port or starboard displacement, or 

 rolling. This is, of course, all without spin. 



If, however, I apply to the solid, as it lies on the 

 tray before me, an impulsive couple with my fingers, 

 so as to make it rotate about one of the minimum 

 diameters (that is, of course, a diameter about which 

 the moment of inertia is a maximum), the solid shows 

 that when spin is applied the equilibrium is unstable. 



The ellipsoid at once sets itself on one end, and then 

 rotates in stable equilibrium with the long axis nearly 

 vertical. This is a very remarkable result. The 

 centre of gravity has been raised, and the equilibrium 

 is now stable. The spin has altered the conditions 

 of equilibrium completely. 



Of course, it was pointed out to us that all these 

 phenomena are weil shown by the ordinary spinning- 

 top, spun by the unwinding from it of a string when 

 the top has been skilfully thrown from the hand. 

 Schoolboys are not encouraged now (indeed they are 

 discouraged by prefects and other important person- 

 ages) to play with tops and marbles, and thus many 

 phenomena of spin and collision which some of us 

 used to observe are missed. The swaying round of 

 the axis of a top when rising just after spin to the 

 " sleeping " position, and the smiilar conical motion 

 of the axis when the top is about to fall, give examples 

 of precessional motion, of, in fact, the astronomical 

 phenomenon called orecession of the equinoxes. 



Precession was illustrated by the interesting old 

 model of a terrestrial globe which I have here (Fig. 2). 

 You see that the globe is weighted so that a pin pro- 

 jecting from the north pole rolls round a ring, that 

 is. a narrow cone fixed in the earth rolls in the inside 

 of a cone fixed in space. These cones have their ver- 

 tices at the earth's centre, the axis of the fixed cone 

 is perpendicular to the ecliptic and its seml-angle is 

 23° 27', that is, an angle equal to the obliquity of the 



