OF ARTS AND SCIENCES. 49 



specific gravity of iron to be 7.2 and the specific heat to be 0.113, 

 we find that through each square centimeter of the inner surface 

 of the cylinder wall exposed to the steam during the icliole time of 

 admission, there have entered 2.4 X 7.2 X 0.113 (= 1.95) c. g. s. 

 units of heat. I estimate the surface exposed to steam in cylin- 

 der and port when admission begins, to be 1700 sq. cm. If we 

 estimate the piston face and port surfaces to be only two thirds as 

 effective for condensation as an equal surface of cylinder head, 

 we may reduce this area to about 1400 sq. cm. The moveinent 

 of the piston during admission exposes about 700 sq. cm. more 

 of the curved cylinder surface. If we count this as being equiv- 

 alent to 350 sq. cm. exposed during the whole of admission, we 

 have 1400 + 350 (= 1750) sq. cm. as the effective area of the 

 surface upon which condensation occurs during admission. The 

 whole amount of heat absorbed through this surface in this time is, 

 then, 1750 X 1.94 (= approximately 3400) c. g. s. units. This 

 would correspond to the condensation, at 49 lbs. absolute pressure, 

 of 3400 -=r 510 (= approximatel}'^ 6.7) grams of steam. The weight 

 of steam in the cylinder at the end of the forward stroke was, ac- 

 cording to the indicator, about 13.5 grams. The weight of steam 

 in the cylinder at cut-off was considerably less than this, probably 

 about 10 grams. According to this calculation, then, the amount 

 of steam condensed during admission was about two thirds as 

 much as the amount not condensed, and about one half of the 

 condensed portion was re-evaporated during expansion. The 

 amount re-evaporated during the exhaust was probably much less, 

 for the curves of Figure 6 indicate that much less heat flowed 

 back to the inner surface during exhaust than during expansion, 

 and we know that a very considerable amount of heat passes through 

 the wall. 



A very rough estimate of water consumption, made May 8th, 

 indicated that about 23.5 grams of water and steam passed through 

 each end of the cylinder at each stroke. The calculations, or con- 

 jectures, just made account for rather less than three fourths of 

 this, but the discussion goes to show that the ebb and flow of heat 

 indicated by these thermo-electric experiments is of the right order 

 of magnitude to account for a large part of the cylinder condensation, 

 and to encourage the hope that a careful and extended set of such 

 experiments would yield results of great certainty and value. I 

 am not prepared to undertake such a labor at present, but the 

 thermopiles which I have used in this preliminary investigation 



VOL. XXVIII. (n. S. XX.) 4 



