228 



Mr. J. Joly. 



rise to an inflow of the surrounding steam. This condition tends to 

 correct what error might arise from radiation from points in the 

 vapour near the surface of the body. Vapour so precipitated will, 

 in fact, be carried by the indraught and thrown upon the substance. 

 In view of this, care is taken so to construct the calorimeter that no 

 sharp cross-draughts play upon the substance on the entry of the 

 steam, which might possibly diminish the protective effect of the 

 indraught, (d.) It is possible that during this first period, if the 

 walls of the calorimeter heat more quickly than the substance, some 

 radiation might occur from the walls to the substance, appearing as 

 a minus error in the result. It is to be added, however, with regard 

 to (d) that the entire duration of this first period is very short, and 

 that the quantitative results of experiments on substances show com- 

 plete independence of surface conditions. 



During the second period, in which the weighing is effected, the 

 substance is to be considered as surrounded by 'a medium which, with- 

 out change of temperature, maintains the inner surface of the walls 

 of the calorimeter uniformly at the temperature of the substance or 

 very nearly so, and which itself acts as a screen very opaque to radia- 

 tion. So that, so soon as the substance has ceased absorbing energy, 

 the conditions are very favourable to preserve it from the effects of 

 further action from radiation. Nevertheless, there is an amount of 

 radiation effect continuing uniformly during this period, and this 

 on a sensitive balance is perceived by continued observation. This 

 was first pointed out to me by Professor Himstedt, of Darmstadt, who, 

 after the appearance of the papers on this method of calorimetry, 

 kindly sent me the results of his observations on instruments of the 

 types described both by Professor Bunsen and by me. He found the 

 weight of the substance hanging in the steam was not absolutely 

 constant, but was subject to an accretion of some 3 or 4 milligrams in 

 an hour. 



As it was desirable to ascertain the cause of this increment and 

 how far it could be reduced, and as, too, it was quite conceivable that 

 occasion might arise when allowance for it would have to be con- 

 sidered, I made a considerable number of protracted observations 

 upon it. It is sufficient to observe here that the effect seems with 

 most probability a radiation effect. It is greater for rough than 

 smooth bodies. Lamp-blacking the substance, or the inside of the 

 calorimeter much increases it. Cooling the outside of the calorimeter, 

 as by spraying cold water upon it, increases it. Calorimeters with 

 double walls, having bright reflecting surfaces, show less increment. 

 For large calorimeters it is less than for small ones. It is uniform or 

 nearly so for however long the experiment is continued. On the other 

 hand, increase of the rate of flow of the steam through the calorimeter 

 does not seem to affect it, which appears to differentiate it from any- 



