PEALE.] TEMPEEATUKE ANALYSES OF THEEMAL WATEES. 387 



periments of M. Donney are interesting in this connection. His results 

 are as follows :* 



1st. The constancy of the boiling j)oint of water depends on its con- 

 taining air in considerable quantity. 



2d. There is a marked difference between the boiling j^oint of "water 

 with air and water without air. 



3d. A small quantity of air dissolved in water is sufficient to attenu- 

 ate greatly the cohesion existing between the molecules of the water. 



4th. Water freed from air as far as possible has its cohesion so in- 

 creased that a higher temi^erature is necessary to evaporate it, and 

 the boiling point is raised. 



M. Donney raised water to 275° F. When water so heated does boil 

 the production of vapor is instantaneous. 



The water of geysers nnd hot springs has been boiled and reboiled 

 for an inconceivable period, and is freed from air as no other water is. 

 In addition, also, the irregularity of the channels offers obstruction to 

 the free escape of the steam and it becomes superheated. In many 

 springs, however, the surface does not boil, and yet tar down in the 

 tube ebullition and the formation of steam do occur, as the throbbings 

 evince. As the bubbles of steam come towards the surface they meet 

 with colder water and condense. In this way the water at the surface 

 gradually becomes hotter and hotter. Spring No. 41 in the Giantess 

 Group in the Upper Fire Hole Basin shows this condensing action. 

 Large bubbles come in at the bottom of the funnel-shaped basin, but 

 soon disappear, drawing a current of water towards the orifice, and the 

 next moment it is repelled by the succeeding bubble. 



CHAPTER VII. 



CHEMISTRY OF THERMAL WATERS. 



By a. C. PEAI.E, M. D., AjSD Hexry Leffmann, M. D. 



ANALYSES OF WATERS. 



In this chapter we shall have reference mainly to the analyses of 

 waters from the Yellowstone National Park. 



When si)eaking of the classification of the springs on a preceding page 

 we mentioned the fact that they differ from ordinary mineral springs only 

 in the smaller proportion of ingredients and in usually haviug a larger 

 amount of silica. We shall therefore consider the analyses with j^rimary 

 reference to the proportion of silica, especially in comparing them with 

 analyses of other waters, with the single exception of the Gardiner's 

 liiver Springs, where the calcareous element predominates. 



None of the muddy waters have as yet been analyzed, but the character 

 of the deposits warrants us in saying that they differ from the others 

 principally in the amount of iron oxide and alumina contained. 



The analyses of waters from the springs of Yellowstone National Park 

 presented here are, we believe, the first approaching completeness that 

 have ever been made. Unfortunately the bottles for securing the specn- 

 mens reached us in the field after many of the most interesting s])rings 

 had been passed. For this reason we have no s])ecimens from the Ui)per 



*Li vol. 17, Memoirs of Eoyal Academy of Sciences and Belles-lettres, Brussels. 



