330 T. A. Jaggar— Conditions affecting Geyser Eruption. 



the " Cliff cauldron' 7 and was not known to be a geyser until 

 1878, when it was first seen in action : it continued to play at 

 irregular intervals for four years and then ceased, resuming its 

 eruptions in 1888. These were continued for only a short sea- 

 son, and since that time the vent has continuously discharged 

 vast volumes of hot water, without explosive eruptions. There 

 is evidently a delicate balance of adjustments between the 

 hydrostatic and thermal forces which govern the actions of 

 this torrent in the depths ; it is possible that the level of the 

 source is affected by exceptional variations in the mean annual 

 rainfall. We should thus expect a period of eruption to fol- 

 low within a year of a season of exceptional dryness, and con- 

 sequent diminution of head.* In the case of Old Faithful 

 regular eruptions have taken place about once an hour ever 

 since its discovery in 1870 : it is a very old vent, as indicated 

 by the seventy feet of sinter which have been built into the 

 walls of its conduit. It is probable that the water-column 

 became confined by reason of the gradual erection of this 

 enclosing shaft ; where there may have been formerly continu- 

 ous overflow, the column is now in equilibrium with its source, 

 and is ejected only under accumulated steam pressure with 

 rhythmic regularity. The tube fills to within a few feet of 

 the orifice soon after each eruption ; this implies percolation 

 into the tube of cooler waters through lateral ducts : as it takes 

 practically a uniform period, some sixty-five minutes, to heat 

 this new column to a state of ebullition at the base, and has 

 done so for twenty-eight years of human record, it is fair to 

 conclude that the heat supply is constant. 



While this coincidence of overflow with irregularity of erup- 

 tion, and confinement with regularity, is in general accord with 

 the facts of observation, it is not essential to the theory that 

 every irregular geyser should visibly overflow at the surface or 

 that every regular one should be enclosed. A non- eruptive 

 spring may have subterranean outlets, and an overflowing one 

 may play at quite regular intervals ; in such a case ebullition in 

 the depths will depend on the ratio of the amount of inflowing 

 cooler water to the amount of heat supplied : in some cases, 

 too, the overflow may be from a separate superficial source. 

 Complications in the conduit system, such as steam-filled cham- 

 bers^ curved ducts and connections with other geysers, will 

 readily account for the composite eruptions observed at some 

 of the vents : a long steam period following the first eruption 

 may be induced in our apparatus by partially reducing the 



* It is noteworthy, though hardly relevant to this case, that in Virginia City, 

 Montana, northwest of the Yellowstone Park, the mean annual rainfall for 1877, 

 the year preceding the first observed eruptions of Excelsior, was much less than 

 the average of the mean annual records for the next four years. 



