56 J. G. Graham— Experiments with an Artificial Geyser. 



down into the iron cylinder, a geyser is formed of practically 

 the same type that Sir George's theory requires. When this 

 is done, this change is at once observable in the eruption. 

 They are unpreceded by steam bubbles. It is evident that this 

 must be so ; for the peculiar form of the tube and chamber 

 are of no avail in Sir George's theory unless the temperature 

 of the water reaches the boiling point before the water-level 

 in the subterranean chamber has been forced down to the 

 opening of the tube into the cavity. And until this level is 

 reached no steam bubbles of any consequence can escape. The 

 conclusion which I would draw would be that those geyers in 

 which the eruptions are markedly preceded by bubbles can- 

 not be classed as " McKenzie's." 



To turn now to the main subject of this paper, that is the 

 effect of soaping geyers. 



It has long been thought that by throwing stones or putting 

 soap into a geyser, a premature eruption could be brought on 

 in the cases of geysers erupting periodically, and that in some 

 cases, hot springs, not known to be active, have by this means 

 been caused to throw out the water in their basins in geyser- 

 spouts. It is evident, however, that this matter is not capable 

 of absolute proof ; for, as none of the geysers are perfect in 

 regard to the equality of their periods, we can never say posi- 

 tively that the geyser would not have erupted when it did if 

 it had not been soaped. Or, in the case of forced eruptions of 

 hot springs, it might be considered that these springs were 

 geysers with periods so long that the nature of the spring was 

 now for the first time revealed. To be sure that such would 

 be the case in many instances is improbable but not impos- 

 sible, especially as but comparatively few springs have been 

 successfully experimented upon. If now, in the artificial 

 geyser, it can be shown that with the same conditions as to 

 the amount of fluid, heat, barometric pressure, etc., the periods of 

 eruption are shorter when the soapy fluid is used than when 

 the fluid is simply water, it will then be conclusively proven 

 that the soap does cause a premature eruption. The question 

 of why it so acts, may then be attacked. 



There are many difficulties in the way of making these 

 seemingly simple observations. 



It is necessaiw that the amount of fluid in the geyser should 

 be the same in all observations ; but owing to splashing 

 and evaporation, this has a constant tendency to vary. The 

 splashing was provided against by filling the geyser not sim- 

 ply to the top of the tube, but part way (3'5 cm ) up the funnel 

 as well, which had the effect of deadening the eruption. The 

 error due to evaporation was provided against by adding a 

 determined amount of fluid at the end of each eruption. By 



