THE AID OF THE ACHROMATIC FRINGES. 53 



I now used this vibrator in connection with the coil-tester (33, primary) 

 obtaining trustworthy results throughout. The data (table 5) obtained for 

 the additional resistance of 9,000 ohms (resistance of circuit, 700 ohms) are 

 shown in the upper curve of figure 56. A few of them (40, B) are low, due to 

 incidental reasons, but the line as a whole indicates a quiet band-width of 

 i scale-part. Hence if the inductive resistance is relatively negligible, 



(R+ 700) (5-1) 



is the effective voltage and should be constant for each coil, increasing from coil 

 to coil with n, the number of turns. The result is shown in the lower curve, 

 figure 56, wh'ch is quite as nearly straight as the inevitable inaccuracies (5 data 

 too small, interrupted induction, irregularities in winding, insufficiencies of the 

 elementary equation) can possibly permit one to expect. The current in B 

 at 9,000 ohms could just be detected in the dynamometer and could not have 

 exceeded 7 X io- 5 ampere, so that 5 = i is again equivalent to less than 5 X io~ 6 

 ampere, as before. A fraction of this was quite observable. 



36. Organ=pipe. It seemed worth while (in connection with the present 

 dynamometer) to ascertain whether the node in an open organ-pipe could be 

 detected by bolometric measurement of the temperature alternations adi- 

 abatically produced at the node of the sounding-pipe. To test this a c-pipe, 

 4 feet long and about a square decimeter in section, giving a frequency of 

 130 per second, was harnessed to an appropriate blower, avoiding overtones. 

 The first bolometer was a grid of fine platinum wire 0.003 cm - m diameter and 

 in all about 48 cm. long, stretched out on a square frame of wood. By aid 

 of a stem attached to the latter, the grid could be lowered at pleasure into the 

 organ-pipe. The bolometer was put in circuit with a storage-cell, an ammeter, 

 a telephone, and a key. The resistance of the grid was about 56 ohms and that 

 of the telephone 87 ohms. 



The bolometer was now supplied with current up to 0.16 ampere, but no 

 note whatever could be obtained from it. At times high-pitched noises (prob- 

 ably of microphonic origin) were heard, but nothing else 



The platinum wire bolometer was now replaced by one of gold leaf, the part 

 between the terminals on the frame being about i cm. long and i cm. broad. 

 Contrary to my expectations, the resistance was here unfavorably low, not 

 exceeding a few ohms as compared with the 87 ohms in the telephone. On 

 lowering the bolometer into the organ-pipe and closing the circuit (currents 

 up to o.i 6 ampere) a loud clatter was always apparent, but no certain evidence 

 of the note c . I heard microphonic noises , often 

 very loud, due to the wind-currents. Cutting 

 down the breadth of the gold foil to a few milli- 

 meters made no difference of consequence. A c= 

 similar experiment was then tried with very 

 thin silver foil with the same negative results. 



A more promising form of the experiment was for the present postponed, 

 because of the noise of the pipe, which would here have to be blown hard to 

 bring out the first overtone. In this case, figure 57, there are two nodes at 



