MEASURING ELECTEICAL CONDUCTIVITY 29 



provided with openings in the center (Fig. 5). This is 

 desirable in many cases and makes it possible to get a 

 high resistance with less tissue. 



Care must be taken to see that liquid does not leak 

 out of the space around the electrodes while the appa- 

 ratus is out of the liquid. If a leak should occur fresh 

 liquid may be added by means of a 

 medicine dropper. With a proper ad- V? 



justment of the rubber disks and suf- 

 ficient tissue to give elasticity no 

 leakage should occur. 



In regard to the accuracy of the 

 readings it may be said at the outset Flo . 5 ._ Hard rubber disks- 



that under favorable conditions sue- SJST*^ tui 



T ,i . T in section). 



cessive readings on the same material 

 do not vary more than 1% from the average. This is as 

 great accuracy as can ordinarily be hoped for in biological 

 work and there is no object in striving to get greater 

 accuracy than this in the apparatus itself. 



It is usually desirable to introduce a variable capaci- 

 tance or an arrangement such as is suggested by Taylor 

 and Curtis (1915), by Taylor and Acree (1916) or by 

 McClendon (1920). In the writer's experiments the 

 capacity of the apparatus, filled with living Laminaria 

 and lifted out of the sea water, was about one thousandth 

 of a microfarad. 



An advantageous arrangement suggested by Profes- 

 sor G. W. Pierce is shown in Fig. 6. 



The frequency is of some importance. The writer 

 has found a thousand cycles convenient; this may be 

 obtained by means of an "audio oscillator" (such as 

 is used in wireless telegraphy) as furnished by the 

 General Radio Co., or by means of a toothed iron wheel re- 



