PREFACE. 



This paper consists of four parts, the first of wliicli contains a determination 

 of the weight of one litre of oxvoen under standard conditions at the sea level in 

 latitude 45°. The second [»art contains a similar determination for hydrogen; hy 

 the use of a method which is new in this application, although previously used by 

 Buff in the case of some other gases, the accidental ei'iors of observation have been 

 made smaller than is usual with so light a gas, while it is lioped that a certain 

 source of constant error also has been eliminated. 



The third part contains au attempt to determine the volumetric composition 

 of water. Both Scott and myself have made determinations of this ratio in which 

 the ffases were measured in eudiometric tubes. But the value thus found was 

 smaller than when, afterwards, considerably larger volumes of the gases were 

 measured in vessels of greater capacity. On account of this unexplained discrep- 

 ancy, three determinations of the volumetric ratio by means of three different 

 processes were thought to be desirable, and the apparatus requii'ed was made 

 ready. Only the least helpful of the three methods, however, has been actually 

 completed. When the other methods were to be used, a Avoikman who had been 

 taught to watch ray apparatus when left in action during my absence fiom the 

 room, had been assigned duties in another building, and the attempt to work with- 

 out this assistance resulted in many accidents, and ended in completely disabling 

 the apparatus; this part of the paper is therefore but fiagraentary. 



The fourth pai't of the papei- contains a series of syntheses of weighed 

 quantities of water from w^eighed quantities of hydrogen and of oxygen. A second 

 series was begun, but the accidents just mentioned put an end to it. Perhaps, 

 however, the series which is here described will have some value even without the 

 confirmation which it was hoped to obtain from an independent series of obsei-va- 



tions of the same kind. 



Excepting one combustion furnace belonging to Adelbert College, all the 

 apparatus used was purchased or constructed expressly for this investigation, 

 which has been carried on in a basement stoi-e-room in one of the college buildings 



