448 A. M. Banta 



to keep the whole contents of the bottle near ii°C. was added 

 from time to time in such amounts as to maintain a nearly constant 

 level in the bottle. Tubes of four sizes were used. The bore of 

 the largest was 118 n in diameter and allowed a flow of water at 

 the rate of 73 cc. per hour. This will be referred to as current 

 No. I. The next in size had a diameter of 90;^, and permitted a 

 flow oi 25 cc. per hour; its current will be designated as No. 2. 

 The third was 38/x in diameter and permitted a flow of 18 cc. per 

 hour; its current will be callled No. 3. The smallest was 26/x in 

 diameter and allowed a flow of 15 cc. her hour; its current is No. 

 4. In a few cases a fifth tube i2/x in diameter and allowing a flow 

 of about 4 cc. per hour was used; current No. 5. The diameters 

 of the tubes were only approximately obtained but the rates of 

 flow of watfr through the tubes were readily and accurately 

 determined and these were made the basis of selecting the cur- 

 rents of various strengths. 



The individuals to be experimented upon were given at least 

 fifteen mmutes to become settled in the dishes before experimenta- 

 tion began. The corresponding parts of an individual of each 

 species were tested in succession by a given current, but no regu- 

 lar sequence was followed in testing the various parts. After 

 all the tests upon a given pair were made with one strength of 

 current, the experiments were repeated with the weaker currents 

 till all four currents had been used. In making these tests, the 

 end of the tube which directed the current was always held under 

 water and at 4 to 5 mm. from the part stimulated. Care was ex- 

 ercised to have the current of water flow squarely upon the part 

 undergoing the test and so directed that other parts of the animal 

 were not aff"ected by it. 



Records were kept after the same plan as was used for the tests 

 with bristles (see p. 441 ), except that in the experiments with cur- 

 rents no records were made other than those of the vigor of the 

 responses, for, even in the experiments with bristles this feature 

 was finally found to be the only significant one. The following 

 record. Table III, in which current 5, as well as the four usual 

 ones, was used, will illustrate the results obtained from stimulation 

 of this sort. 



