948 



Transactioxs of the American Institute. 



in the other class of instances, accounts for the discrepancy. In 

 most of the diagrams, the piston of the indicator vibrated very 

 violently, being projected upward by the sudden influx of steam 

 to a point on the scale above even the boiler j^ressure, thence return- 

 ing partially and vibrating above and below its true position during 

 the entire stroke, which it must be recollected occupied but half a 

 second. I^o effort was made to check these vibrations, as all the 

 ordinary expedients for the purpose involve a loss of accuracy in 

 the results, either by increasing the friction of the indicator or by 

 preventing the full effect of the steam thereon. In some instances, 

 however, the time of the several changes in pressure would conspire 

 to increase the vibration to an extravagant degree, and in others the 

 same influence would nearly or quite destroy it. This latter effect 

 is very finely apparent in the production of the diagram shown 

 above, which is a VQ^vma^faGeimile of the third diagram in the above 

 table, one in which the steam pressure was nearly up to the highest 

 point allowed, and the load extremely light, — that of the lever 

 alone, which exerted a force on the break strap equal to about 400 



pounds. In this figure, A A is the atmospheric line drawn by the 

 indicator, B B the imaginary line for absolute vacuum, C C a similar 

 line showing the boiler pressure, and D the point thereon where the 

 steam should have commenced to expand, in order to sink to the 

 same pressure at the termination of the stroke, according to the 



