Clarice and Schneider — Experiments, etc. 303 



its reflection, as done at Bretenil, leaves nothing to be desired ; 

 but I could not work with large enough tubes, and errors of 

 setting were not my principal errors. 



Measured in this way, my three best measurements gave for 

 the vapor tension : 



Difference of TJ and I in 200ths mm . 



Before admitting acid. After.' Difference in mm . 



31 33 +-01 



9 11 +-01 



— 12 —13 — -005 



This result is considered to be purely negative, as errors due 

 to the change of the height of the barometer during the experi- 

 ment and those introduced by irregularities in the capillarity 

 of the mercury (in spite of shaking the tube) would be sufficient 

 to account for the measured tension. A larger tube, would 

 have avoided some of the trouble, but would have increased 

 very much the labor of manipulation. 



In several other experiments, attention was carefully fixed 

 upon the top of the mercury column in the tube b just as the 

 acid rose through the other tube (a), in order to observe any 

 instantaneous movement of the column. The movement, if 

 any, was very slight — certainly not in excess of the above 

 measurements. I therefore conclude that the tension is not 

 greater than about - 01 mm . 



Aet. XLL — Experiments upon the Constitution of the Nat- 

 ural Silicates ; by F. W. Clarke and E. A. Schneider. 



During the past six years the constitution of the natural 

 silicates has received a good deal of attention in the laboratory 

 of the United States Geological Survey. A number of papers 

 have been published by one of us, partly theoretical and partly 

 analytical ; but the evidence so far considered has been drawn 

 only from careful analyses of various minerals, and a study of 

 their obvious relations, their associations, and their alteration 

 products. Within certain limits the results obtained have been 

 satisfactory and suggestive ; but in several cases it was found 

 that ordinary analysis failed to discriminate between possible 

 alternative formulae ; and it was plain that the uncertainties 

 could only be cleared up by new lines of experimental investi- 

 gation. With such investigations the present paper has to do. 

 Sixteen minerals, including varieties, all of the magnesian 

 group, have been examined ; and it has been found quite pos- 

 sible to separate some of them into distinct fractions of definite 



