CORRELATION BETWEEN METABOLIC GRADIENTS. 327 



table represent divisions of the scale of the galvanometer; where 

 the differences were small, movements of the indicator of less than 

 one division are recorded as .5. Freshly collected material was 

 nearly always used for the galvanometer tests. 



Direct current for the galvanotaxis tests was obtained from a 

 number of cells in series or from a no-volt direct-current gen- 

 erator. Metal electrodes were used, owing to the high resistance 

 of non-polarizable electrodes. The voltage was regulated with a 

 variable resistance shunted across the line in such a way that any 

 voltage from zero to the maximum could be obtained simply by 

 moving a sliding contact. When testing an animal, the slide was 

 placed at the zero end of the resistance, the animal introduced into 

 the testing dish, and the potential gradually increased to a strength 

 which would evoke a definite response. Strong currents or pro- 

 longed exposure to the current often alters the galvanotactic re- 

 sponse. The records kept were not sufficiently complete to enable 

 us to calculate the current density in the galvanotaxis experiments 

 with any degree of accuracy, as we did not regard such data as 

 significant for the purpose of this investigation. 



III. SPONGES. 



The work on sponges recorded here was done by Hyman at 

 Woods Hole in the summer of 1919. * 



i. Metabolic Gradients. Potassium permanganate was used for 

 the study of the metabolic differences along the axis of two simple 

 sponges, Grantia and Leucosolenia. Potassium permanganate is 

 reduced by protoplasm, giving a brown stain, and the rate of 

 appearance of this stain, as well as the depth of its color, is a 

 rough measure of metabolic differences, since regions of higher 

 metabolic rate reduce the permanganate more rapidly (see further 

 Child, '190). In applying this method to simple sponges it is 

 preferable, owing to the thickness of the animal, to make a median 

 longitudinal section and watch the appearance of the brown color 

 along the cut surface. This was done in the case of Grantia, 

 while entire individuals were tested in the case of Leucosolenia. 

 In many individuals of Leucosolenia and Grantia which were tested 

 in potassium permanganate a slight but evident gradation of stain- 

 ing power was noted along the axis, the oral or oscular end staining 



