10 



MR. B. F. CUMMINGS ON 



downhill when I tested them in the spring, but in the autumn 

 this preference changed and they went up. In work with 

 an inclined plane, Molge aHstata was chiefly used. The Newts 

 were kept in water in large earthenware pans under similar 

 conditions so far as possible. The observations were made in 

 the early morning or at night, always in subdued light. The 

 plane was of wood and measured 125 centimetres by 82 centi- 

 metres 5 millimetres. A line was drawn transversely across the 

 plane, and each Newt on removal from the pan was allowed 

 to crawl on a level svirface adjoining until manreuvred with the 

 finger into the right direction, when it was swiftly lifted on to the 

 incline and left in a horizontal position. In this way I cut out 

 errors arising from compensatory head-movements. An interval 

 of ten minutes was allowed each animal before being tried again. 

 This prevented errors arising from simple habit formation, which 

 in the Newt is rapid. 



Table I. shows the results of experiments performed in the 

 spring when the Newts, used singly, walked downhill in 80 per 

 cent, of the tests. Similar i-esults were obtained by using Newts 

 in series ; twenty Newts were used. The number of positive re- 

 sponses was in this way reduced to 67 per cent., which perhaps 

 is nearer the truth. Table II. was obtained in the autumn 

 when the Newts were endeavouring to leave the water for winter 

 quarters. The Newts were then negative and walked uphill 

 72 per cent, of the times. In this table the Newts were used in 

 series, the first row being worked by twenty Palmates and the 

 rest by twenty Crested. 



Table I. 



