30 



KNOWLEDGE. 



[February 1, 1894. 



of the extinct tliree-toed horses known as the hipparions ; 

 the middle toe being very large, while the two lateral ones 

 were small and functionless. There is, moreover, some 

 reason to believe that in one genus the toes were reduced 

 to a single large one on each foot, as in the modern horse. 

 Be this as it may, the fact that there existed in South 

 America a group of ungulates which exactly paralleled the 

 horses in the evolution and structure of their feet is one 

 of the most wonderful features in mammalian development 

 that has ever come under our notice, and serves to render 

 these prototheres (as the group is called) worthy of the 

 best attention of the student of parallelism. It may be 

 added that the high specialization of these animals is of 

 itself suiiicient to dispose of the idea that the strata in 

 which their remains are entombed are equivalent in age 

 to the lower Eocene of Europe. 



This completes our survey of the groups of extinct 

 ungulates iseculiar to South America, which, brief as it is, 

 serves to show how interesting they are alike to the pure 

 anatomist and to the evolutionist, and also indicates their 

 importance from a distributional point of view. With 

 regard to the latter, it may be observed that since all these 

 animals are clearly more or less intimately related to the 

 ancestors of the odd-toed ungulates of the Old World and 

 North America, while the edentates of South America 

 have no near relatives elsewhere, and the extinct mar- 

 supials of the same area appear akin to those now living in 

 Australia, it would seem that at some remote epoch there 

 must have been a brief connection between North and 

 South America, during which one or more primitive 

 ungulates obtained entrance into the latter area, where 

 they subsequently multiplied and developed into the 

 numerous forms above noticed. If this be true, it will be 

 evident that South American ungulates were originally 

 immigrants from the north ; those which are absolutely 

 peculiar to the country having originated from an incursion 

 which took place early in the Tertiary period, while the 

 later types nearly akm to those of North America and the 

 Old World did not arrive till the late Pliocene epoch, when 

 a second connection must have been established between 

 the two continents. 



WEIGHING THE EARTH. 



By J. J. Stewart, B.A., B.Sc. 



SINCE Newton's splendid verification of the theory of 

 universal gravitation, several attempts have been 

 made to determine the mass of our planet, and 

 hence- the mean density of the materials com- 

 posing it. ■ Amongst the earliest of these which 

 approached accuracy was the experiment ■ by Henry 

 Cavendish, the eminent philosopher, who anticipated 

 last century some of Faraday's remarkable electrical 

 discoveries. The method he employed was suggested by 

 the Rev. .John Michell, and is generally known as the 

 Cavendish experiment. ^Michell constructed an apparatus 

 suitable for the operation, but does not seem himself to 

 have made any experiments : and alterations and improve- 

 ments were made in the apparatus by Cavendish. The 

 main principle of the method is that of comparing the 

 attraction exerted by a leaden sphere on a small ball with 

 the attraction exerted by the earth on the same ball — that 

 is, with the weight of the latter. In the apparatus of 

 Michell and Cavendish the attraction between two spheres 

 of moderate size is balanced by the elasticity of a wire ; 

 the amount of torsion in the wire required to resist the 

 attraction of the spheres being found, this gives the force 



of attraction which is equal to it. This force is then 

 compared with the attraction of the earth on one of the 

 spheres — that is, with the weight of the sphere ; and hence, 

 as the diameter and therefore the volume of the earth is 

 known, its mass is deduced. 



Michell's original apparatus consisted of a wooden arm 

 six feet long, made so as to unite great strength with little 

 weight. This arm being suspended in a horizontal 

 position by a fine wire, from each end of it was himg a 

 leaden ball about two inches in diameter, and the whole 

 was enclosed in a wooden case to protect it from draughts 

 of air. The force required to turn the arm is no more 

 than that necessary to twist the suspending fibre, and as 

 this was very slender, even a very small force, such as that 

 produced by the attraction of a leaden weight a few inches 

 in diameter, is sufScient to produce a sensible displace- 

 ment of the arm. As the force with which the balls are 

 attracted is excessively minute, not more than one fifty- 

 millionth of their weight, a very small disturbing force is 

 sufficient to destroy the success of the experiment. 

 Cavendish found that the disturbing force most difficult to 

 guard against was that arising from variations of 

 temperature. Suppose that one side of the containing 

 case is warmer than the other, this will cause the air next 

 it to be rarefied, and it will consequently rise, while the 

 air on the cooler side will descend, and thus currents 

 will be produced which interfere with the deflections of 

 the instrument to be observed. 



In the apparatus as modified by Cavendish the balls 

 were suspended by wires from a light arm made of deal 

 and strengthened by a silver wire, so as to have the form 

 of an open girder ; thus lightness and strength were com- 

 bined. The apparatus was placed in a room kept constantly 

 closed, and the motion of the arm was observed from the 

 outside by means of a telescope ; thus errors, which might 

 arise from the approach of the observer causing changes in 

 the temperature, were avoided. The leaden weights, which 

 were spheres a few inches in diameter, could be moved 

 from the outside by means of a cord passing over a pulley. 

 The weights were placed close to the balls, so that they 

 tended to twist the suspending fibre round in the same 

 direction and cause a movement of the arm. They were 

 then turned round and placed on the opposite sides of the 

 respective balls, so that they tended to move the arm 

 round in the opposite direction. Between these observa- 

 tions of the deflections produced, the line joining the 

 weights was set at right angles to the arm, so that they 

 exerted no attraction on the balls, or rather, equal and 

 opposite attractions which neutralized each other. The 

 position of the arm could be read off on ivory scales 

 graduated to one-twentieth of an inch, and read by a 

 vernier to one-hundredth of an inch, and estimated to 

 less. The scales were lit up by lamps placed outside 

 the chamber, and their light was thrown by a lens on 

 to the scales, the chamber being otherwise dark. 



When the arm was drawn aside it vibrated backwards 

 and forwards, the vibrations lasting a long time ; thus the 

 position that would be occupied by the arm if it were at 

 rest had to be calculated. This was done by observing 

 three consecutive extreme positions of the arm, the first 

 and third being towards the same side and the second 

 one in the opposite direction. The intermediate observa- 

 tion was compared with the mean of the first and third ; 

 this gave the position the deflected arm would have 

 occupied if it had come to rest, damping of the vibrations 

 being allowed for. 



Next, the time of vibration was measured ; from this 

 could be deduced the force acting on the suspended balls. 

 Possible error due to magnetic efi'ects was looked for, but 



