454 Prof. Tyndall on the Action of Free Molecules on 



its absorbent power was restored. Then, with dry air as a 

 carrier, I introduced smoke into the experimental tube, until 

 it far exceeded in density that suspended in the London air 

 when Magnus drew my attention to it. The quantity of heat 

 intercepted by this smoke proved to be only a fraction of that 

 absorbed by the perfectly invisible aqueous vapour. 



On his return to Berlin he resumed his labours. He had 

 been especially impressed by the experiments with the open 

 tube; and to this point he directed his chief attention. " The 

 result of this experiment," he writes, " was so surprising, and 

 so little in accord with what I had found by other methods, 

 that on reaching home I determined to repeat the experi- 

 ment." He did so, with this result: — " I have," he says, 

 " repeated the blowing-in of dry air and moist air many 

 hundred times; but in no single case was the deflection such 

 as to indicate a greater absorption by moist air "*. Humid 

 air, in his experiments, produced the deflection of heat ; dry 

 air the deflection of cold — a result diametrically opposed to 

 mine. In London he had seen that my deflections were as 

 large as I had affirmed them to be ; but he had not criticised 

 them with the view of ascertaining whether they were, or were 

 not, in the right direction. In these new experiments, how- 

 ever, he had, he thought, hit upon their origin. The moving 

 air had reached the face of the thermopile, producing, 

 when humid, heat by condensation, and, when dry, cold by 

 evaporation. 



I read the account of these experiments with some concern; 

 for it was thereby made plain to me that Magnus had by no 

 means realized the anxious care that I had bestowed upon my 

 work. The testimony of an independent observer would, I 

 thought, set the matter right. My apparatus, carefully ad- 

 justed, was accordingly handed over to Dr. Frankland, who 

 minutely tested every point involved in, or arising out of, 

 the objection of Magnus. He verified all my results. His 

 opinion as to the accuracy of the method of compensation is 

 worth recording. " In conclusion," he writes, " I cannot 

 but express my surprise and admiration at the precision and 

 sharpness of the indications of your apparatus. Without 

 having actually worked with it, I should not have thought it 

 possible to obtain these qualities in so high a degree in deter- 

 minations of such extreme delicacy " f. To this may be 



* Poggendorff's Annalen, 18G3, vol. cxviii. p. 580 ; Phil. Mag. 1863, 

 toI. xxvi. p. 2-3. 



t The total heat here employed amounted to 86°-2 of a quadrant. 

 This exceedingly large deflection was neutralized by the radiation fruin 

 the compensatiug cube. But so accurately were the two sources balanced, 

 and so constant was the radiation on both sides, that the determinations 

 were made with ease, and without sensible disturbance or fluctuation. 



