June 4, 1886.] 



SCIENCE. 



501 



from the orifice, the intensity of the vibratory- 

 changes propagated to it varies with this distance, 

 as for the jet itself. It is simply necessary, then, 

 to insert into the nappe two platinum electrodes 

 in circuit with a telephone and a battery having 

 an electromotive force of from twelve to thirty 

 volts, to get an accurate and faithful reproduction 

 of the jet-vibrations, Loud sounds can thus be 

 obtained from a jet which is finer than the finest 

 needle, and the arrangement constitutes a highly 

 sensitive ' transmitter ' (figs. 7 and 8). 



A jet-transmitter, in its simplest form, consists 

 essentially of a glass jet-tube which is rigidly at- 

 tached to a sound-board, and supplied from an 

 elevated reservoir containing some conducting- 

 liquid (distilled water acidified with one three- 

 hundredth of its volume of pure sulphuric acid is 

 the best), and a couple of platinum electrodes 

 embedded in an insulator, such as ebonite, 

 against which the jet strikes. The jet may 

 issue from a circular orifice, about 0.25 mm. in 

 diameter, in the blunt and thin-sided end of a 

 small glass tube. Much smaller jets may be used, 

 but, for one of the given size, the pressure re- 

 quired for distinct transmission of all kinds of 

 sounds will not exceed thirty inches. The receiv- 

 ing-surface is the rounded end of an ebonite rod, 

 through the centre of which passes a platinum 

 wire. The upper end of the rod should be about 

 1 mm. in diameter, , and should be surrounded by 

 a little tube of platinum ; and the end of the cen- 

 tral wire and the upper margin of the tube should 

 form a continuous slightly convex surface with 

 the ebonite, free from irregularities. The inner 

 and outer platinum electrodes are joined respec- 

 tively to the terminals of the circuit. The jet is 

 allowed to strike on the end of the central wire, 

 and, thence radiating in the form of a nappe, 

 comes into contact with the tube, thus completing 

 the circuit. Tne dimensions of the apparatus may 

 be varied to suit jets of different sizes ; it is highly 

 desirable, however, that the jet nappe should well 

 overlap the inner margin of the ring-shaped elec- 

 trode. 



With small jets the impact disturbances are so 

 feeble, that slight precautions are necessary to in- 

 sulate the receiving-surface from the orifice, un- 

 less the former is placed low down in the path. 

 The strength of battery may be increased until the 

 escape of electrolytic gas-bubbles causes a faint 

 hissing noise in the telephone. The liquid, on its 

 way to the jet, should pass downwards tlirough a 

 wide tube lightly packed with coarse clean cotton, 

 by which minute air-bubbles which violently dis- 

 turb the jet, and small particles of dust which 

 might obstruct the orifice, are stopped. This tube 

 should never be allowed to empty itself. 



Experiments are given to show that in this in- 

 strument the jet may act upon the electric current 

 in two ways : first, by interposing a constantly 

 changing liquid resistance between the electrodes ; 

 and, second, by causing changes in the so-called 

 ' polarization ' of the electrodes. In one form of 

 instrument, namely, that in which both jet and 

 electrodes are entirely immersed in a mass of liquid 

 of the same kind as the jet liquid, the action must 

 be entirely at the surface of the electrodes. 



In the latter case a liquid jet becomes similar in 

 structure and properties to a jet of air in air, and 

 the velocity at different points when it is steady 

 and when it is disturbed varies in precisely the 

 manner already described. 



The author briefly passed in review the leading 

 facts to be accounted for, and laid stress upon the 

 parallelism of the properties of gaseous and liquid 

 jets. Some shadow - photographs of vibrating 

 smoke jets have shown that these also present 

 drop-like swellings and contractions which grow 

 along the jet-path. The most satisfactory expla- 

 nation of the phenomena will then be one which 

 refers the vibratory changes in jets of both kinds 

 to the same origin. 



The beautiful and well-known experiments of 

 Plateau have supplied a satisfactory explanation 

 of the normal vibrations of a liquid jet in air. He 

 has shown that a stationary liquid cylinder, whose 

 length exceeds a certain multiple of its diameter, 

 must break up, under the influence of the ' forces 

 of figure,' into shorter cylinders of definite length, 

 which, when liberated, tend to contract into drops. 

 Now, the jet being regarded as such a stationary 

 cylinder, we have a satisfactory explanation of the 

 musical tone resulting when its discontinuous part 

 strikes upon a stretched membrane, and when the 

 impact disturbances may be in any way conducted 

 back to the orifice. These disturbances then accel- 

 erate the division of the jet after it leaves the ori- 

 fice. Plateau endeavored to show that division of 

 the jet might take place at other than the normal 

 points, thus explaining Savart's conclusion that 

 a jet can vibrate in sympathy with a limited range 

 of tones. Lord Rayleigh, moreover, has recently 

 shown that the inferior limit of this range is not 

 so sharply defined theoretically as Savart's experi- 

 ments would prove it to be. 



Both Savart and Magnus, however, describe ex- 

 periments in which a water- jet, carefully protected 

 from impact and other disturbances, does not 

 exhibit the peculiar appearances characteristic of 

 rhythmical division ; and the author's experiments 

 conclusively prove that this rhythmical division 

 does not take place in a well-insulated jet. While 

 the tendency so to divide may therefore be admit- 

 ted, and the normal rate of vibration of the jet 



