Although Watt was a prolific writer when communicating with business partners and friends, he did not write any scientific papers to be published.

In a letter to Roebuck on 24 May 1768 Watt wrote "I speak within bounds when I say that every cubic foot of the contents of the cylinder will require only one cubic inch of water to be evaporated" [1]. As 12 inch make a foot Watt, with simple means had already evaluated that one unit of water transform into 12³ = 1728 units of steam. That was close to real figure of 1 liter of water at 4°C gives 1673 liter of steam of 100°C.

Heat capacity of steam

According to his account, published in Robinson's book, James Watt conducted several experiments in February and March 1781 to determine the amount of latent heat. For this purpose, he had fixed a copper pipe steam-tight to the spout of a water kettle. He let the steam condense in a pan of water and measured the increase in temperature as well the increase in the weight of the water. Watt deducted to which temperature water would have been heated by the same amount of heat. At that time, heat was considered the same as temperature, so the latent heat was expressed as a temperature.

Watt summarized the results of his experiments in a table:



Watt dropped the 2nd and 11th measurements that were too far off and calculated the latent heat as 949.9°F. If we calculate the total heat, i.e. the sensible heat for heating water from freezing point to boiling point plus the temperature equivalent for the latent heat, we arrive at an average of 1161.9°F. As the boiling temperature of water is 212°F we can deduct from James Watt's experiments that the proportion of total heat to sensible heat is 949.9°F/212°F = 5.48. If we compare that with the internal energy u for water at boiling point = 419.6 kJ/kg and the internal energy of steam at 100°C and at atmospheric pressure = 2506 kJ/kg we arrive at a proportion of 419.6/2506 = 5.98. Watt was only (5.48-5.98)/5.98 = 8% off the actual figure. This is an excellent result for the primitive circumstances he had conducted his experiments.

Pressure- temperature relationship of steam

Already in the winter 1764/1765 James Watt conducted experiments to determine the steam pressure at a few temperatures above boiling point. In 1773/1774 Watt resumed his experiments, now for temperatures below the boiling point of water. The latter data is shown in the table below.




[1] source cited in H.W. Dickinson & R. Jenkins, "James Watt and the Steam Engine" first published 1929, as the "Doldowlod Papers"