174 E. ABBAY ON THE BUILDING-UP OE THE 



like small columns in the colder water d (fig. 3), but only reach, 

 up to the under surface of the hotter stratum m — a formation which 

 only occurs in the upper' basins, where the water is hottest. 



Fig. 3. — A Sinter Basin. 



Showing columnar elevations of the floccuient deposit c. 



A reason for the existence of these columns is, I think, suggested 

 by the analogy of the rain-clouds. It is well known that when two 

 masses of air saturated with aqueous vapour at different tempe- 

 ratures intermingle, there is always a surplus amount of vapour over 

 and above what is required for the complete saturation of the re- 

 sultant mixture. This surplus vapour condenses and takes the form 

 of a rain-cloud. In the same way, I think, when water saturated 

 with silica at different temperatures, coming from two different 

 basins, intermingles, an analogous process to condensation may come 

 into operation, and a surplus precipitate of silica be formed in places 

 where this intermingling chiefly takes place. If eddies are caused 

 where the two currents intermingle, they may explain the occurrence 

 of these curious little columns, which stand perfectly isolated in some 

 of the basins. 



As an instance of the effect of temperature in causing terrace- 

 basins rather than fiat terraces to be formed, it may be mentioned 

 that, besides the facts already stated in regard to the lowest steps of 

 the main terrace, which has been described, on the opposite side of 

 the lake, at a distance of half a mile, is a second system of terraces, 

 called the " Pink Terrace," in all about 80 ft. in height. The water 

 issuing from the geyser belonging to this terrace is considerably 

 below boiling-point ; and, as a consequence, only a few of the topmost 

 steps are in the form of basins, all the middle and lower ones being 

 flat, *and showing on the outer walls only the stalactitic forms 

 already spoken of in connexion with the White Terrace, and not 

 the peculiar structure of the miniature basins. 



It will be evident that the processes of building up the terrace- 

 basins cannot go on long without some general effect being produced 

 on the geyser itself ; for the tendency is manifestly for the geyser 

 to build itself in and thereby to be forced into and up the hill. 

 Several facts appear to confirm the view that such changes have 

 actually taken place. In the first place, the geyser-funnel is not 

 situated on the natural slope of the hill, as one would expect if the 

 geyser had broken out at its present elevation above the lake, but 

 it is situated at some distance w T ithin the hill, as shown in fig. 4. 



