ADDENDA. 



Ix the time that has elapsed between the completion of the 

 manuscript and the printing- of this book, some improvements 

 in the foregoing' methods have been effected, as follows. 



1 (page 66). The chief drawback to the use of the Pfeffer cell is the 

 uncertainty of its working. However, the porous cups and filter bougies, 

 having a straight tube sealed into them with sealing-wax, always work 

 well with me up to a pressure of somewhat over one atmosphere, above 

 which the molasses (which I generally use) escapes through the cup into 

 the outer vessel. It is, by the way, somewhat better in preparing the 

 cups to place them in the copper chloride or sulphate solution and 

 exhaust thoroughly under an air-pump. Also, it is very easy to make 

 the diffusion-shells semi-permeable by thoroughly soaking and \varming 

 them in water until all air is removed, when they are to be filled with the 

 potassium ferrocyanide solution and placed overnight in copper sul- 

 phate solution. They may then be filled with molasses and will always 

 quickly give a pressure of over an atmosphere, after which the mem- 

 branes appear to burst, for fine jets may be seen issuing from the sides 

 of the shell. The shell thus prepared will raise a column of water three 

 or four times as high as when unprepared. 



Perhaps this subject is not worth the trouble given to it; but its 

 value lies in the excellence of its illustration that the work of the plant 

 can be done by ordinary physical forces. 



2 (page 68). For example, the extreme vapor tension for 15 C. is 

 12.70 mm. of mercury, that is, about -^ of an atmosphere; for 20 C. it is 

 17.39 rnm., or about ^ of an atmosphere. The error in the gauges will 

 of course be in the direction of a lesser result; that is, owingto the vapor 

 tension, the pressure is really greater than it appears to be from the 

 level of the liquid. 



It may, however, be exactly calculated from the following table, in 

 which the pressure is expressed in millimeters of mercury for each de- 

 gree of temperature. To apply it, one of course pays no attention to it at 

 the reading made when the tube is first sealed, but in all subsequent read- 

 ings one first calculates the pressure in fractions of atmospheres (or else 



i37 



