Sept. 4, 



1879] 



NATURE 



447 



gauge. Kain striking upon the inner and opposite side of the 

 gauge would be thrown out, and so an equilibrium rim would be 

 conj-tructed, as the gain on one side would be balanced by the 

 loss on the other side. 



With this view, the author had an 8-inch gauge made and 

 tested alongside of an 8-inch Giaisher gauge. The sloping lip 

 of the Giaisher gauge had an angle of 45° from the perpendicular, 

 and the rim of the equilibrium gauge was "8 in. deep, 'iS in. 

 in thickne.'s, sloping off on both sides at an angle of 3° from the 

 perpendicular. Both gauges were fixed at Croydon, 4 feet above 

 the ground, and 259 feet above Ordnance datum. These gauges 

 had been working side by side for 55 1 days, from January 5, 

 1S78, to July 5, 1879, during which period rain or snow has 

 fallen upon 306 occasions. Upon 43 occasions it was found that 

 the rain collected in the Giaisher gauge exceeded, by a small 

 amount, the rain in the equilibrium rim-gauge, and on two oc- 

 casions the quantity in the new gauge exceeded that in the 

 Glai>her gauge. Upon 261 occasions the rain in both gauges was 

 absolutely equal. On all occasions, it should be observed, the 

 rain from 'both gauges was invariably measured in the tame 

 graduated measuring glass. On the 45 occasions when the 

 Giaisher gauge collected most rain, the wind without exception 

 was high. On the two occasions when the equilibrium rim- 

 gauge collected more rain than the Giaisher gauge, it was pro- 

 bably due to dew, the equilibrium gauge presenting a larger 

 surface for condensation than the other gauge. As the Giaisher 

 gauge was not calculated to, contain snow, all falls of snow are 

 recorded in the equilibrium rim-gauge, which is constructed to 

 hold about one foot in depth of snow. 



The total quantity of rain collected in the Giaisher gauge 

 during the period of observation, plus the snow as caught in the 

 equilibrium rim-gauge, was 46"68 in., and the quantity collected 

 in the equilibrium rim-gauge was 45"45 m., showing a difference 

 of but half per cent. In all probability, however, the small 

 excess measured by the Giaisher gauge would tend to compensate 

 for the losses by evaporation in periods of small rainfall and at 

 other times, and therefore, as a measuring 'gauge, the Giaisher 

 pattern of gauge, when tested by a gauge of the description 

 mentioned, gives results in practice which may be taken as 

 correct. 



Summary of Results 



On the Kelardation of Phase of Vibrations transmitted hy the 

 Te/epAone, by I'rof. S. P. Thompson. — It was predicted 

 from theoretical considerations by Dubois-Reymond that a dif- 

 ference of phase, amounting to a quarter of a complete vibration, 



would be found to exist between the diaphragms of two associated 

 Bell telephones, the receiving telephone being a quarter of a 

 vibration behind the transmitter. A more complete theory, 

 worked out independently by Helmholtz and Weber, gave a 

 somewhat contradictory result, and required only a small 

 difference of phase. Recently Konig in a series of delicate 

 experiments, effected an optical comparison by the method of 

 lissajous of the vibrations of a pair of telephones, replacing the 

 vibrating discs by tuning-forks armed with mirrors. The 

 experiment is a delicate one, and is performed under con- 

 dition not free from objection. The author has proposed the 

 following method of observing. A pair of Bell telephones are 

 suspended by wires of about a metre in length, so as to oscillate 

 as pendulums, to frames so disposed as to avoid thcpossibility of 

 any mechanical transmission of the vibrations. Below the point 

 of rest of each telephone, and at some little distance from it in 

 the plane of its swinging, is placed a steel magnet. After the 

 lengths of the wires have been so adjusted that the telephones 

 will swing in identical periods, one telephone is set swinging. 

 As it alternately approaches and recedes from the magnet, the 

 induced currents traversing the second telephone set it swinging. 

 In every case the difference of phase observed amounted to one 

 quarter. 



In the case of those telephones which transmit vibrations by 

 varying the resistance of the circuit, instead of varying the 

 electromotive force, there is no such retardation of phase pro- 

 duced in the ordinary electromagnetic receiver. If, however, 

 the current so transmitted is first passed through an induction 

 coil, a retardation of phase of one quarter is produced, and in 

 the case of several successive inductions the retardation amounts 

 to an additional quarter for every additional induction. This 

 remark applies only to vibrations of harmonic and quasi- 

 harmonic type. Vowel sounds, which consist of compound 

 harmonic vibrations, are unchanged to the perception of the 

 single ear, which is unable to distinguish differences of phase, 

 or between compound sounds which differ from one another only 

 in the difference of phase of their components. The vibrations 

 of consonantal sounds, 'on the contrary, depart more and more 

 widely from their original type at each successive induction. 



In the case of Edison's motographic or electro-chemical 

 receiver, the velocities, not the displacement of the disc, are 

 proportional to the strength of the currents received. Hence 

 vibrations already retarded one quarter in transmission, as is 

 the case with those of the carbon transmitter in conjunction 

 with its induction coil, always used with this instrument, are 

 restored to their primitive phase. The vibrations of this receiver 

 therefore agree in type, not with the vibrations of the induction 

 current (which correspond to the derived function of those of 

 the original vibration), but with those corresponding to the 

 function of which the vibrations of tbe induction current are 

 the derivate ; that is to say, they agree in type with the 

 primitive vibrations of whatever form. Hence in the receiving 

 telephone of Edison consonantal sounds which depart widely from 

 the purely harmonic type are better rendered than in a telephone 

 which like that of Bell both retards the vibrations in phase 

 and alters them in type. 



On some New Instruments recently constructed for the Continua- 

 tion of Researches on Specific Inductive Capacity, by J. E. H. 

 Gordon, B. A. — Mr. Gordon exhibited and explained the fol- 

 lowing new instruments which he has arranged during the last 

 year : — 



1. A miniature five-plate induction-balance, similar in principle 

 to the large balance exhibited at the Dublin meeting, but intended 

 for the examination of crystals and other precious substances 

 which cannot be obtained in sufficiently large quantities for the 

 large balance. 



The large balance requires the dielectric plates to be 7 inches 

 squire and 4 '° S 'w^t^ thick. For the small balance it is 

 sufficient to make them 2 inches square and ^ inch thick. 



2. A gauge for measuring the thickness of the dielectric plates 

 to Tiiiinr incli- 



3. A new form of quadrant electrometer for use with the small 

 induction balance. 



The capacity of the smaller plates of the little induction 

 balance is so minute that when they are attached to the quadrants 

 of the electrometer of ordinary construction (Elliott pattern) 

 disturbances in them produce hardly any effect on the needle, on 

 account of the much greater capacity of the quadrants of the 

 electrometer. 



In order to construct an electrometer whoce quadrants should 



