NEW RESEARCHES ON LIQUID AIR. 137 



is left iii the vessel, the influx of heat is diminished to one-sixth part of 

 the amount entering without the metallic coating. The total effect of 

 the high vacuum and silvering is to reduce the ingoing heat to one- 

 thirtieth part, or, roughly, 3£ per cent. Vessels constructed with 

 three dry-air spaces only reduced the influx of heat to 35 per cent. An 

 ordinary mercury vacuum vessel is therefore ten times more economical 

 for storing liquid air, apart from considerations of manipulation, than 

 a triple annular-spaced air vessel. It has been suggested that the 

 metallic coating of mercury does no good, because Pictet has found 

 that all kinds of matter become transparent to beat at low tempera- 

 tures. The results above mentioned dispose of this assumption, and 

 direct experiment proves that no increase in the transparency of glass 

 to thermal radiation is effected by cooling it to the boiling point of air. 1 

 An ocular demonstration of the correctness of the above statements 

 can easily be shown by mounting on the same stem three similar 

 double- walled test tubes, two of which have been simultaneously 

 exhausted and sealed off from the air pump together, while the third is 

 left full of air. One of the vacuum test tubes is coated with silver in the 

 interior. The apparatus is shown in fig. 1, Plate III. A Las the annu- 

 lar space filled with air; B and C are exhausted, C being coated with 

 silver. On filling liquid ethylene to the same height into each vessel 

 and inserting corks with similar gas jets and igniting the escaping gas, 

 the relative volumes of the flames is roughly proportional to the influx 

 of heat, and resembles what is shown in the drawing. It is satisfac- 

 tory to have independent corroboration of the advantages of the use of 

 vacuum vessels, and this may be found in a paper by Prof. Kamerlingh 

 Onnes, of Leyden, communicated to the Amsterdam Academy of Sci- 

 ences, 1896, entitled "Beinarks on the liquefaction of hydrogen, on 

 thermodynamical similarity, and in the use of vacuum vessels," in 

 which he says: "In the same degree as it becomes of more importance 

 to effectuate adiabatic processes at very low temperatures, the impor- 

 tance of the vacuum vessels of Dewar will increase. It seems to me 

 that they are the most important addition since 1883 to the appliances 

 for low-temperature research." . . . "It is a rejoicing prospect that 

 practical engineers will doubtless feel the want of such nonconducting 

 mantles. For as soon as this stage is reached numbers of heads and 



1 At a meeting of the French Academy in 1895 a paper hy M. Solvay, of Brussels, 

 was read, in which my 1892 device of vacuum vessels was attributed to M. Cailletet, 

 and tacitly accepted hy him. In 1875 I had already used a highly exhaustive vessel, 

 of similar shape to the vacuum test tube, in calorimetric experiments. See paper 

 on "The physical constants of hydrogenium," Trans. Roy. Soc. Ed., Vol. XXVII. 

 Even as late as April, 1896, Professor Tilden, D. Sc. F. R. S. of the Royal College of 

 Science, in a paper entitled " L'Appareil du Dr. Hampson pour la liquefaction de 

 Fair et des gas," communicated to the Revue G6nerale des Sciences, thought proper 

 to write as follows : " Un manchon de verrc, dans lequel on a fait le vide (mauchon 

 semblable a ceux d6crits par Cailletet ou Dewar)." Where did Professor Tilden find 

 Cailletet's description of a vacuum vessel? This is not the only statement in the 

 paper requiring correction. 



