ABSORPTION 



i4S 



brane and inversely proportional to the size of the tube, it is desirable that 

 the membrane be made as large as possible, either by greatly enlarging the 

 bottom of the tube, or else by using the membrane in the form of a cup or 

 bag. Several forms of good osmoscopes are illustrated in the accompany- 

 ing figure, all adapted from ordinary laboratory materials. Upon the whole 

 the most convenient is that shown by the letter F, using a Schleicher and 

 Schuell diffusion shell of 16 mm. diameter (a suitable supporting glass ring 

 for which is supplied with my normal apparatus, page 46). In all of these 

 forms it is essential that the connection between membrane and glass be 

 perfectly tight, though this is not at all difficult to accomplish if the mem- 

 brane is first soaked in water (five to ten minutes) and is then tied tightly 

 in place with strong well-waxed thread. In the construction of osmoscopes 

 regard must be given to convenience of inserting the solution, which is readily 



37 



w 



A B C D E F 



Fig. 38. — Various forms of adapted osmoscopes; X%. 



Double lines are glass, black lines are membranes, and dotted areas are stoppers. 



effected in the stoppered forms, but may be accomplished, though with some 

 trouble, in the thistle-tube form by pouring the liquid cautiously down a 

 wire. A somewhat well-known make-shift instrument, more troublesome, 

 however, to prepare, as I should suppose, than some of the above, is that 

 using the membrane of an egg, described by Bergen in his elementary 

 text-books. 



An osmoscope of this kind may very readily be converted into an osmom- 

 eter by using a graduated tube and a known quantity of solution. Further, 

 it would be possible to make any of these forms autographic by use of a 

 frictionless float and a recording wheel like those later described under 

 Growth. Where an experiment is long continued, the molasses or other 

 solution may be kept from fermentation by the addition of 2-3% of forma- 

 lin, while evaporation may be checked by a covering of oil. 



