OF ARTS AND SCIENCES. 47 



the engine, and noting at frequent intervals the number of strokes 

 per minute. The general type of the diagrams is shown in Figure 5. 

 The maximum pressure during admission was generally near 34 lb. 

 above atmospheric pressure, and cut-off occurred near one quarter 

 stroke. Wishing to have the diagrams affected as little as possible 

 by the action of the governor, I depended largely upon the load, 

 which was a dynamo with closed circuit, to regulate the speed of 

 the engine to sixty strokes per minute. The steam pressure, how- 

 ever, was by no means constant, and the speed sometimes rose or 

 fell two or three strokes from the desired figure. This inconstancy, 

 and the discrepancy of thermometers, one or two of which were 

 broken during the experiments, may help to account for some lack 

 of consistency in the results obtained, which are presented in tabu- 

 lar form on page 46, estimated stem corrections having been 

 applied to the thermometer readings. 



I have omitted from this table all results of observations made 

 previous to April 17, 1891, for the reason that in tins.' earlier 

 observations the time during which contact lasted was about three 

 times as long as in later ones. Nevertheless, the results obtained 

 with the long time of contact are in general agreement with those 

 obtained later, and as in the table there is no record for the depth 

 0.101 cm. at that part of the stroke marked by K o\\ the indicator 

 diagram, I shall fill this gap, as well as I can, from a comparison of 

 the earlier observations with the later, which comparison indie 

 the temperature 123°. I shall undertake to show by means of a 

 diagram, Figure 6, the temperature condition of the inner part of 

 the cylinder wall as indicated by my observations at three points of 

 the stroke, C, E, and .ST Distances into the cylinder wall from its 

 inner surface are laid off along the base line of the diagram from 

 right to left. Temperatures in excess of 110°. 5. the temperature 

 of the outer surface of the cylinder wall, are measured upward from 

 this base line. Line K indicates the condition of things jus* before 

 admission of new steam; line C, the condition at or mar the time 

 of cut-off; line E, the condition near the end of the forward stroke. 

 The broken parts of curves are extensions beyond the region of 

 actual observation. It is easy to see that an error of our or two 

 degrees in the observation of temperature at the depth 0.05J cm. 

 would make a great difference in the position of the broken part oi 

 the curve to which this observation belongs, and it is evidenl thai 

 the broken part of curve C cannot be correct, for it indicati 

 the inner surface of the cylinder wall a temperature Borne degi 



