SEVENTEENTH ANNUAL MEETING. 93 



III. 



Upon the super-saturation of vapors there exist, so far as I know, no published re- 

 searches; and aside from Professor Thomson's paper (already cited) and a brief allusion 

 to thai paper in Maxwell's* Theory of Heat, the subject has attracted but little atten- 

 tion. My attempts to realize the conditions under which vapors might be cooled below 

 the boiling point without condensation were begun in 1880. Believing the immediate 

 cause of condensation to be the presence of nuclei of the substance in question in Liquid 

 form, f I collected steam in inverted glass tubes tilled with heated oil ; the mouths of 

 the tubes dipped into an oil bath, and in some cases the whole tube was submerged. 

 The steam, however, condensed upon the sides of these tubes, in the form of dew, when- 

 ever the bath was allowed to cool to the boiling point of water. The film of oil upon 

 the glass was doubtless insufficient to prevent the steam from coming in contact with the 

 hygroscopic surface of the glass. 



Some experiments made in 1882 were more successful. A glass tube (fig. 2) 5 cm. 

 in diameter and 50 cm. long, open at both ends and having two tubulatures through 

 which thermometers were inserted, was placed with its lower end in a bath of mercury 

 and tilled with linseed oil. The mercury was heated to 150° and a few drops of water 

 were introduced into the column of oil. Upon reaching the hot mercury at the bottom 

 of the tube the water boiled, and bubbles of steam rose through the oil without condens- 

 ing until a point was reached which had a temperature of 70°, when it condensed sud- 

 denly. During the course of the experiment, the oil being gradually heated from below, 

 this 70° level rose and the point of condensation rose with it until the entire column of 

 oil was hotter than 70°, when the steam condensed at the surface. Upon substituting 

 bi-sulphide of carbon for the water in this experiment, and water for the oil, a similar 

 result was reached. The bi-sulphide of carbon vapor rose uncondensed through water 

 whenever the latter was above 35°, a temperature 10° below the boiling point of the sub- 

 stance. Steam generated outside the apparatus and introduced into the column of oil 

 from below behaved in the same manner as that produced by boiling drops of water in 

 the tube itself. When led, for comparison, into warm water, it condensed explosively, in 

 the usual manner, until the water itself had begun to boil. 



To determine whether other liquids than oil would serve as an envelope for the iso- 

 lation and consequent super-saturation of steam, I introduced dry steam into moderately 

 warm dry mercury. The behavior of the vapor within the mercury could of course 

 not be directly observed, but the surface of the latter became coated with minute globules 

 of water in the form of dew, indicating, I think, that it had risen through the mercury 

 in the form of vapor and condensed at the surface. Whenever condensation below the 

 surface occurred, as in the cases in which wet steam was used, no dew appeared and the 

 water collected in considerable quantities before rising to the surface of the mercury. 



I have recently repeated these experiments in a slightly modified form. A glass 

 tube (fig. 3) rather more than a metre in length, two centimetres wide, and closed at the 

 lower end, was filled with oil, and immersed to a depth of four or five centimetres in an 

 oil bath at 200°. Such a column of oil heats slowly, and the temperature decreases 



* Clerk .Maxwell, Theory of Heat, p. 124. 



t The view of the processes of ebullition and condensation upon which this opinion is based lias 

 been more clearly stated in Maxwell's treatise (just referred to) than in any work with which I am 

 acquainted. According to this view tin y are special cases of the general process known as fluid diffu- 

 sion. Ebullition consists of an inrush of liquid particles into gaseous nuclei contained within the body 

 of the liquid or occluded upon the walls of the vessel. The absence of such nuclei precludes boiling 

 aiel permits of super-heating. Condi nsation is the reverse process in which particles of a vapor enter 

 liquid nuclei, becoming part of them and producing drops of the liquid, 'the absence of such nuclei 

 might be expected from the nature of the case to prevent condensation and make super-saturation 

 possible. 



