384 PROCEEDINGS OF THE AMERICAN ACADEMY. 



condition of negative response would under similar circumstances 

 turn toward the smaller light. For this reason, although it would 

 not matter in testing the reactions of the animal to the two lights, 

 whether it were positive or negative, it is exceedingly important that 

 the insect should be in one or other of these states and remain so 

 without change throughout the experiment. It is so much easier to 

 keep Ranatra in the positive condition than in the negative one, 

 and the insect reacts so much more definitely and decidedly when 

 positive, that, so far as possible, only individuals in this conditioii 

 were used. A number of specimens were taken from the water and 

 placed on the table. Most of them immediately went into the death 

 feint, as described by Holmes ; but there were usually a few that re- 

 mained active. Some of these were occasionally positive, though more 

 often they were negative and began crawling away from the light. 

 After a time the others began gradually to come out of their death 

 feint, and these, in nearly all cases, were positive. They were al- 

 lowed to crawl toward the light until they became fully active and 

 could be picked up and handled without fear of changing their re- 

 actions. That individual was then selected for experimentation which 

 appeared most strongly and persistently positive. If, as often hap- 

 pened, they reached such a state of excitement that they attempted 

 to fly to the light, the method adopted by Holmes to prevent this 

 was resorted to, namely, fastening down their wings with asphalt 

 varnish. 



No regular separate series of experiments were made to test the 

 reactions of Ranatra to the lights singly, but trials of this kind were 

 made from time to time in the course of the tests with both lights by 

 screening off one or the other, and exposing the animal to the remain- 

 ing light. Furthermore, the method used for bringing the animals into 

 normal orientation at the beginning of each trial aftbrded a means of 

 observing immediately any change in the character of the reactions. 

 The preliminary orientation was accomplished in the following manner: 

 Two runways were constructed by tacking strips of pasteboard, painted 

 black, to the sides of wooden blocks on each of which was mounted 

 a small incandescent lamp (Figure 14, /, /') registered as 2 c. p. By 

 means of a single switch, either of these lamps could be lighted at will 

 simply by throwing the switch handle one way or the other. When 

 the switch was midway, both lamps were thrown out of circuit. If it is 

 now desired to test an animal to the action of two lights, Lg and Sm, 

 having it first headed north, so that Lg is to its left and ^Sm to its 

 right, the animal is picked up by its breathing tube and dropped into 

 the runway So with its head pointing toward No. The switch has pre- 



