112 REPORT OF THE PSYCHOLOGY COMMITTEE 



well as a horizontal motion for training. This was done 

 optically by introducing a total reflection prism into the eye 

 pieces of two of the skeleton guns. To reproduce for the 

 gun-pointer the errors of his training mate, so that he could 

 get practice in watching the vertical wire while firing, an 

 artificial training error was introduced mechanically at fixed 

 points of the motion of the target platform. At first thought 

 this would seem to give the training error such a mechanical 

 constancy that it could be predicted and therefore ignored. 

 This is absolutely not the case. It follows the non-predict- 

 able character of the main displacement of the target, be- 

 cause the arbitrary training error will appear at every phase 

 of the more or less extended movements of the targets, while 

 on the records they will appear only in homologous parts of 

 the curves. 



"Firing records are introduced in the pointing records by 

 pressing an electric button, which closes a circuit operating a 

 small electric bell, only instead of hitting the bell, the buzzing 

 armature hits the pencil holder. This makes a dot on the 

 records when the gun is standing still, and a waved line in- 

 stead of a simple line when the gun is moving. The firing 

 will be perfect only when these dots and waves appear on a 

 straight line along the theoretically perfect pointing line, at 

 those parts of the record that are free from the artificial train- 

 ing errors. 



"The artificial training error is produced by placing irreg- 

 ularities on the track of the target platform, over which a 

 friction wheel glides. This is attached to a series of levers 

 that add a vertical motion to the main horizontal movements 

 of the targets. Since the optical system transforms the hori- 

 zontal motion into a vertical one for the pointer, it transforms 

 these slight vertical movements of the target into errors of 

 training. 



"The actual uses of the pointing instrument will probably 

 vary according to the needs and inclinations of the groups for 

 which they are available. Quite tentatively, I suggested the 

 following: Entirely raw recruits should probably be intro- 

 duced only to the training mechanisms, without recording all 



