20 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 8l 



They call the acute angle between the two planes a and the acute one 

 between the track of the caterpillar and the horizontal plane 0. The 

 angle a was changed for each series of tests, and in order to have an 

 average angle 6 a caterpillar was repeatedly tested, first with one side 

 and then the other side downward. The results obtained are quantita- 

 tively described in terms of trigonometry. The summary in part is 

 about as follows : The geotactic orientation of tent caterpillars while 

 creeping upon a surface inclined at angle a to the horizontal is such 

 that the path makes an average angle 6 upward on the plane, of a 

 magnitude proportional to log sin a. This is attributed to the fluctua- 

 tion of the pull of the head region upon the lateral musculature of 

 the upper side during the side-to-side swinging brought about in 

 creeping. 



A review of the literature by Crozier and Stier (7) shows that 

 neither the mechanical theory nor the statolith theory is sufficient to 

 explain geotaxis, because neither accounts " for the quantitative rela- 

 tionships between gravitational pull and the amplitude of orientation. 

 There is left the appeal to the proprioceptive results of muscle ten- 

 sions, already suggested to account for certain features of geotropism 

 among insects and molluscs." 



2. GEOTACTIC EXPERIMENTS ON CODLING-MOTH LARVAE 



Recently hatched codling-moth larvae do not seem to be influenced 

 by gravity either in the laboratory or on the foliage of apple trees. 

 They avoid bright sunshine as much as possible, and if there are no 

 interfering factors they crawl in all directions, as if hunting for food. 

 Seventy-five larvae belonging to the fourth, fifth, and sixth instars 

 were subjected to phototactic and geotactic tests in the laboratory. 

 A branch of an apple tree, 30 inches long and bearing leaves, a small 

 apple, and small twigs, was suspended from a chandelier. At the 

 extreme top and bottom of the branch twine was wound loosely around 

 the twigs to furnish a cocooning place for the larvae being tested. 

 After receiving the phototactic test a larva was laid horizontally in 

 one of the forks of the branch, and in such a position the light was not 

 an interfering factor. The results obtained follow. 



The light reactions were found to be a crude index for judging 

 the responses to gravity. Those individuals which were weakly photo- 

 negative or were indififerent to light were generally not yet ready to 

 make cocoons and consequently were not geopositive ; but when ready 

 to spin, or later, they were nearly always strongly photonegative and 

 geopositive. When larvae ready to spin were put on the branch, they 



