144 B. p. UVAROV. 



Dociostaiirus maroccanus, Thvmbg., both of which I have studied for several years. 

 It is true that when the vegetation is ver\' rich, the rapidity and extent of movement 

 of larval swarms is less than in the case of a sparse vegetation, but larvae will move, 

 however densely overgrown with their most favourable food-plants the place may 

 be. Moreover, in both the species referred to, the larvae do not feed during move- 

 ment, save in some exceptional cases, as, for instance, when a swarm has previously 

 wandered for a long time over barren ground and then comes across a field of corn 

 or other rich vegetation. My observations, though far from being complete, leave 

 no doubt that a factor of utmost importance in the movement of larval swarms 

 is temperature, and the following outline of the daily regime of a swarm will help 

 to explain this. 



It is a well-known fact that larval swamis do not move at all during the night, 

 which they pass on plants in a semi-comatose state, caused doubtless by the low 

 temperature. The first rays of the rising sun bring the larvae back to active life, 

 and they begin at once to feed. The hotter the temperature grows, the more active 

 become the insects, and soon they one by one jump or crawl down to the ground, 

 where thev continue to move about, now crawling, now jumping, as if feeling uneasy. 

 Each larva in doing so disturbs its fellows, which leads to still more rapid general 

 movement ; this is at first entirely irregular, but sooner or later, through mutual 

 influence, a common direction of movement is found and the swarm begins its day's 

 march. So far as our observations in the Northern Caucasus go, the average 

 temperature causing the first movements of the larvae lies near 13-15° C. In 

 ordinarv circumstances, i.e., provided the sun shines all day and no unusual changes 

 of weather occur, swamis keep moving the whole day, and as a rule do not feed 

 when they move.* In the afternoon, when the temperature begins to fall, the 

 velocitv of the movement decreases until, in the evening, the swarm stops ; it is 

 ^■ery interesting to note that this evening cessation of movement occurs when the 

 temperature is again near 13-15° C. One b}^ one the larvae crawl up plants and 

 begin to take their evening meal. There is no doubt that swarms do not choose 

 their resting places, but simph' stop where they are overtaken by the critical 

 temperature. I ha\-e seen many swanns which have stopped for the night at entirely 

 barren spots just after they have crossed a strip with abundant vegetation, or only 

 a few yards before reaching it. 



Such is the behaviour of larval swarms of migratoria under ordinary conditions 

 of weather, and all my observations lead me to the definite conclusion that their 

 movement has nothing to do with hunger and depends entirely on thermotropism 

 (probably negative), and on another, as yet little known, tropism which is displayed 

 by the tendency of each larva to repeat the movements of its nearest fellows and to 

 move in the same direction as they do.t 



The behaviour of larval swarms under abnonnal conditions of weather lends 

 further support to this conclusion. If the day is colder than usual and the tempera- 

 ture does not rise above 15^ C, the swarm remains and feeds all da}' where it 

 spent the night. If the temperature falls during a normally hot day, the swarm 

 stops long before its usual time. Especially interesting and instructive is the 

 following example : if a swarm is moving at the temperature in sunshine not much 

 above the critical point and the sun is temporai-ily hidden behind a cloud, so that 

 the temperature falls below that point, the swarm stops, or in an}' case its movement 



* I intentionally avoid discussing here the question of the direction of movements and its 

 causes, since it is very complicated, and besides, it has no direct connection with the chief problem 

 we are now investigating ; one point, however, is worth mentioning : that the direction of move- 

 ment has, as a rule, nothing to do with the looking for food, since, for one thing, larvae do not 

 feed during the day, and they often move from fertile spots into entirely barren places. 



t This kind of tropism is not uncommon among other insects, and even vertebrates ; a herd 

 of cattle will afford excellent examples of it. 



