Paper Atmometers i^or Studies in Evaporation. 285 



of the thimble, howevei , and proved unsatisfactory for other 

 reasons. A better contact between paper and tube was attained 

 by the construction of a s])ecial seat for the thimble, of metallic 

 tin. While the latter might be made of glass the tin is more 

 readily manipulated and its extremely slight solubility in dis- 

 tilled water renders it quite satisfactory. The seat is constructed 

 as follows: 



The rounded portion (less than 1 cm.) of the end of a paper 

 thimble is cut off and the little paper cup thus obtained is sup- 

 ported, concave side uppermost, in a mass of sand, care being 

 taken not to alter the original form of the thimble end. A piece 

 of "block' ' tin tubing (about 6mm. ciitside diameter and 10 cm. 

 or more in length) is placed and supported upright in the center 

 of the little paper matrix thus prepared, and the latter is filled 

 with molten tin. After cooling the casting is trimmed and 

 smoothed with knife and file, the bore of the tube is reopened, 

 and the seat is finished. Where a number of inr.truments are to 

 be prepared, better procedures than the one above described may 

 readily be devised; this one has proved satisfactory and can be 

 followed in any laboratory. 



The tin tube is now attached to the burette, in place of the 

 glass of the Piche-Cantoni arrangement, a paper cylinder is 

 passed down over the tin casting and carefully pressed into po- 

 sition, where it should fit snugly (see Fig. 3). W'ith the finger 

 pressing the paper firmly against the opening in the seat, water 

 is now made to rise in the tube, expelling all air before it, until 

 the closed end of the papei cylinder is thoroughly wetted, when 

 the finger is removed. The instrument as thus set up, operates 

 several days in the gieenhouse; frequent filling of the burette 

 should pi event the occurence of a pressure of more than a few 

 centimeters of water column. 



The instrument is completed and rendered stable enough 

 for almost any sort of woik by two additions, that of an arrange- 

 ment to hold the cylinder more fiimly in place and prevent the 

 entrance of air, and that of a support at its lower free margin. 

 My clamp was made as follows. A test tube is selected, into 

 which the filter thimble will just fit, its closed end is cut off (by 

 means of an emery wheel) and ground down till it forms a small, 



