IIAUMONOC.KAI'M TKAC I \( ;S Kl-C OUDIJ) 

 HLECTROlA-rKAIJA'. 



By J. II. \ INCIINT. M.A.. D.Si .. A.R.C.Sc, AM. C. W. JlDi:. l'..S< . 



I.oiuliiii County Council Padilinfiton Tcclinicdl Institute. 



Anv form I iiarmono^rapli may he used in tlicsc 

 expcrimrnls. The ordinary tubular writing pen is 

 replaced 1)\- a steel needle with its jraint carefulh' 

 blunted on a hone so as to have a smooth semi- 

 circular end. This end forms one of the electrodes. 

 The other electrode is a sheet of brass w hich is fixed 

 firmly to the writing table of the instrument. The 

 electrolyte is a solution of ammonium nitrate and 

 (lotassium ferro-cyanide in water. The paper uj^on 

 which the record is to be taken, is soaked in this 

 solution and is then partially dried. The needle is 

 insulated from the harmonograph by being thrust 

 through a cork which fits tlie holder designed to 

 receive the ordinary tubular pen : it is connected by 

 a very thin wire to one terminal of the source of 

 electric current, while the brass plate upon which 

 the prepared paper is |>laced is similarly connected 

 to the other terminal. 



If direct current is eniployeil an electromotive 

 force of about twentv volts \\ ill be found suitable, 

 though this will depend on the kind of paper used, 

 and other circumstances. The provision of a 

 rheostat for regulating the current will be found 

 convenient. When the needle is safely started 

 describing its path on the paper the current is 

 turned on, and is switched off after any desired stage 

 of progress of the record has been attained. Bv 

 interrupting the current by a clock signal the period 

 of vibration of either pendulum can be determined, 

 or if these periods be known an irUerrupted trace will 

 serve the purpose of a chronograph record. The 

 direction in which the trace is niade by the st\le can 

 also be recorded by this method. For exam|)le, a 

 short interruption followed by a longer one would 

 serve the purpose of an arrow-head on the trace and 

 would indicate the relative direction of motion of the 

 style and paper. 



If alternatiiig current is available, pictures like 

 those in the accompanying plate may readily be 

 jiroduced. In this case, the current being reversed 

 man\' times a second, the electricity is flowing in the 

 direction required to leave a trace, only for a small 

 fraction of a second at a time. The trace then is 

 built up of a series of more or less elongated dots, 

 the intensity of each being greatest in the middle, 

 and shading off to nothing at each end. The number 

 of the dots marked per second is nearly constant, 



and is e(iual to the fre(]ucncy of the alternating 

 supply. It follow s from this that the dots are siiaced 

 on the trace so as to show b\' their remoteness or 

 proximity to each other the cpiickness or slowness of 

 the relative motion of the needle and paper. The 

 velocity of the style with respect to the paper is given 

 in centimetres a second b\- dividing the frequency of 

 the suppl\- by the number of dots in a centimetre. 

 It will be found that an effective voltage of about 

 forty is suitable for use with paper of the thickness 

 and quality of good unlined foolscap : but again it 

 w ill be an advantage to have means of varving the 

 voltage so as to obtain the best effects. 



.As with ordinary harmonograph curves, the more 

 simple ratios of period give the best pictures. 

 Figures i-v are all given by the ratio 2:1. the 

 tuning being accurate in Figures iv and v. Figures 

 vi-ix are given by the ratio 1:1, the only figure of 

 these in which unison is exact being Figure vi. The 

 initial phase is maintained throughout the diagram 

 ill Figures i, iii, iv, v and vi, this being very accur- 

 ately so in Figures iv, v and vi. In Figure ii, how- 

 ever, the phase difference of the two pendulums is 

 originally similar to that of Figure iv, and finally is 

 like that of Figure i. Figure vi is the onlv diagram 

 in which the dots are equally spaced throughout 

 one complete oscillation of both pendulums. The 

 velocity in each turn of the spiral is practically 

 constant, but decreases as we approach the centre of 

 the figure. In Figures vii, viii and ix we commence 

 with an approximatelv circular path, but owing to 

 unequal damping of the two motions, and to change 

 (jf phase-difference due to inaccurate tuning, the 

 final curve is a much flattened ellip.se of different 

 orientation from the original one. 



In Figures i ami iii it will be seen that the trace 

 was stopped when the left to right motion had 

 almost died away, while the up and down motion 

 was nearly as vigorous as at first. This effect, 

 which ma\-, of course, be shown with any harmono- 

 graph, whether the printing be electrolytic or other- 

 wise, can be produced either bv using pendulums 

 with bobs of very unequal weight, or by damping 

 the motion of one by means of a flat flexible spring 

 rubbing lightly against the rod of the pendulum 

 so as to press on the rod in a direction at right 

 angles to its motion. 



