GRAPHIC METHOD OF RECORDING MAZE-REACTIONS 55 



tablet having been properly oriented, the experimenter may at 

 any moment tear off a record sheet and continue his tracing 

 almost uninterruptedly. This method may be made to work 

 satisfactorily even when a printed diagram of the maze appears 

 on each record sheet, for the orientation of the block may be kept 

 constant. When the image falls upon the record sheet from below 

 it is necessary to use rather thin paper and to have the drawing 

 board so arranged that the sheets fit neatly and may be quickly 

 placed in position. Although we have tried only methods I 

 and II in this laboratory, we are inclined to believe that it is more 

 satisfactory on the whole to have the image come from below 

 the record sheet. Especially in the first trials with a given 

 animal, the time required is likely to be long, and the experimenter 

 should be able to make his observations without undue discomfort 

 or fatigue. 



We recommend that as soon as the experimental . device has 

 been selected and properly adjusted, a zinc etching, which 

 exactly reproduces the image of the maze as it falls upon the 

 drawing board, be made, and that this be used in the prepara- 

 tion of blank record sheets. For, although a diagram of the 

 maze is not absolutely essential, it has considerable value in 

 connection with the early trials and sometimes prevents errors 

 in the reading of records of later trials. It is extremely laborious 

 and wasteful of time to draw the diagrams by hand, and if 

 hundreds or thousands of record sheets are to be used, the cost 

 of a zinc etching and of printing the sheets will be trivial in 

 comparison with the value of the experimenter's time. 



As appears from the above discussion, we are not in a position 

 to recommend any one of the four methods over the others 

 without careful consideration of the type of maze which is to be 

 used, of the character of the lens, and the location of the ap- 

 paratus. But we are fully convinced that all investigations 

 with the maze should make use of some graphic method of record- 

 ing reactions. There can be no doubt that the data previously 

 obtained from maze experiments have only a fraction of the value 

 which they should have, and that because of the inaccuracy and 

 incompleteness of the records. 



