1] METHODS 127 



account of the strength and continuance of their currents. In 

 other cases, CLARK or DANIELL elements, if enough of them 

 are united in series, will meet the requirements. The character 

 of the electrodes, next, will depend upon the nature of the in- 

 vestigation. Nonpolarizable ones of hair (camel's-hair brush), 

 clay, or paper (plug of filter paper in glass tubing drawn out 

 to a cone) are usually employed, but all of these offer consider- 

 able resistance. The troughs will vary in form and size with 

 the organisms to be contained in them ; some of them will be 

 described in connection with the experiments in which they have 

 been employed. They are all rectangular enclosures having 

 clay ends when it is desirable that these should be iionpolariz- 

 able. For large troughs, sheet-zinc electrodes are used, cover- 

 ing the smaller sides of the trough. Although some of the 

 reflecting galvanometers are more sensitive, a " millammeter " 

 such as that made by the WESTON Electrical Works is a much 

 more convenient instrument and sensitive enough for most 

 work of this sort. The rheochord is practically a low-resistance 

 box, capable of indefinitely fine gradations. This is introduced 

 into the short branch of a divided circuit, so that by varying 

 its resistance a varying share of the current shall be forced into 

 the longer circuit. A very simple and excellent device for 

 altering the strength of current is the " Compression-rheostat " 

 of BLASIUS and SCHWEIZER ('93). This consists of a piece of 

 rubber tubing filled with zinc sulphate, stopped at the ends 

 and introduced into the circuit. By means of a thumbscrew 

 the walls of the middle of the tube may be pressed together, the 

 lumen correspondingly reduced, and the resistance increased.. 

 The induction apparatus usually employed is one invented by 

 DU BOIS-REYMOND. Iii this the secondary coil may be with- 

 drawn from the primary coil to any desired distance, thereby 

 diminishing the intensity of the induced current. Through 

 the action of such an instrument the current is alternately 

 made and broken, and each electrode becomes in quick suc- 

 cession anode and kathode. Since alternating currents cannot 

 be measured by an ordinary galvanometer, an electrometer 

 must be employed. So much concerning apparatus. 



A word should be said about the method of stating the cur- 

 rent employed. Very many authors have been satisfied with 



