162. ELIZABETH LOCKWOOD THOMPSON 



found that this current caused the snails to draw backward 

 into their shells for a little way. They usually resumed their 

 normal movements in a short time. Rarely they retracted so 

 much that they lost their hold on the glass tubing and dropped 

 to the bottom of the tank. The short arm of the Y from the 

 point of divergence to within a millimeter or so of the platinum 

 wires was roughened by means of a band of block tin, 2 mm. 

 in width and 1 mm. thick, passing around it in a spiral. The 

 metal was then insulated with a coat of liquid celluloid and 

 further roughened by a coat of coarse cutting sand imbedded 

 in the celluloid before it hardened. These precautions appeared 

 desirable because it seemed possible that enough mucus might 

 be secreted by the snail in crawling over sand alone, to fill in 

 the spaces between the grains and thus make a smooth surface 

 upon which to travel. This hardly seemed possible over the 

 ridges of metal used, inasmuch as they were far enough separated 

 so that the snail had to sink down into the space between, in 

 passing over them. If the snail turned toward the short arm 

 •at the top of the stem it immediately came in contact with the 

 rough ridges and a little later with the bare electric wires, at 

 which point the punishment was administered by closing the 

 'circuit. No attempt was made to keep the snail on one side 

 ■of the apparatus. It sometimes crawled in an irregular spiral 

 but frequently kept on one side. The whole apparatus was 

 made shorter than the celluloid labyrinth previously used, in 

 ^order to shorten the time used in each test and thus make a 

 'greater number of tests possible in a given time. The method 

 of manipulation and cleaning the apparatus was the same as 

 in the previous tests, and in order to avoid any possible tendency 

 •of the snail to follow the same path in consecutive trips, the 

 labyrinth was frequently turned so that the short arm was as 

 ■often toward the left as the right. The " map " method before 

 described was adopted in making the records. Four snails were 

 used and a total of 664 trials given. The curves (figs. 19-22, 

 Plate VII) were based upon the percentage of errors for each 

 snail per day. 



