Prof. L. T. More on Electrostrktion. 17 



tops of the tubes were kept in their proper positions by 

 wedojes. Tliis was accomplished by inserting three little 

 wedges of wood, or, preferably, of glass, between the experi- 

 mental tube and each of the armatures at the top. Lest the 

 friction might lessen the full effect of the expansion, new 

 determinations for the modulus of elasticity were made and 

 gave the same values recorded in the previous table for that 

 experiment. 



With the wedges in place, a potential of 12 millimetre 

 spark produced some lateral bending, but not nearly so much 

 and the motion was uniform. It was accompanied by an 

 apparent elongation of 60 divisions. That this elongation 

 was due to the side-wise motion was proved by having an 

 assistant press gently against the top of the tube in a horizontal 

 direction until the lateral motion was the same as before — 

 the mirror aoain indicated an elonoation of 60 divisions. 



It was noted, during the experiments with adherent arma- 

 tures, that after a tube had been charged a few times the image 

 moved steadily to indicate a gradual elongation of the tube. 

 This motion was ascribed to the effect of heat. Experiment 

 shows this effect to be more pronounced with the present 

 arrangement. 



The room temperature being constant, a series of charges 

 were given the condenser by exciting the generator con- 

 tinuously until a number of sparks passed, after which it was 

 short-circuited and the image observed ; then the deflexion 

 was recorded at successive intervals of time. 



Spark'le7igtJi, 12 mm. — Number of Sparks, 10. 



Deflexion, I'o minutes after tube was discharged ... 50 div. 

 4-0 „ later 100 „ 



5-5 „ 150 



>) 



Spark-length, 12 mm. — Number of additional Sparks, 10. 



Deflexion, 1*5 minutes after discharge 130 div. 



„ 4-0 „ later 225 „ 



5-0 „ „ 245 „ 



5-0 „ „ 285 „ 



16-5 „ 965 „ 



The outer tube broke with further charging. It was not 

 altogether satisfactory, as it was not accurately cylindrical and 

 straight and was unnecessarily large. It was replaced by a 



Phil. Mag. S. 6. Yoh 6. No. 31. Juli/ 1903. C 



