A REVISION OF THE GENUS LOCUSTA. L. 145 



becomes slower and less defined ; as soon as the cloud has passed and the sun shines 

 again, the swarms starts its march afresh ; even small clouds hiding the sun just 

 for a few minutes produce the same effect. 



There is hkewise an upper limit of temperature that causes the cessation of 

 movement ; and on verj' hot, close days swarms often stop their march during 

 midday ; in that case, however, the larvae do not climb plants to feed but collect 

 in close clusters under the plants, evidently hiding from the direct rays of the sun. 

 This being a more rare phenomenon, I have no precise data as to the exact tem- 

 perature at which it occurs, but anyhow it gives additional support to my theory 

 that the movements of larval swarms depend primarily on thermotropism.* 



After the final moult the wanderings of swarms on foot naturally cease, thoUvgh 

 when the majority of a swarm is in the last larval stage and only single individuals 

 become winged, the latter may often be seen crawling and jumping with the swarm. 



A few days after the last moult, newly-winged insects are incapable of long flight, 

 their elytra and wings being not yet hard enough. This period is passed by swarms 

 in the same spot, devouring vast quantities of food, and it is a very favourable 

 (and also the last) opportunity for their destruction by spraying. 



When locusts are fit for flight, single individuals begin to take wing and fly for 

 a short distance, often circling above the still sitting swarm. WTienever a locust 

 flies near enough to another that is at rest, the latter is disturbed and often takes 

 wing and flies in the same direction ; this is again a manifestation of the same tropism 

 which causes the movement of larval swarms. The larger the number of individuals 

 with fully developed wings, the more often do they take these short flights, and 

 the more other locusts join them, disturbing yet others when circling above the 

 swarm. It is easy to understand that this must necessarily result, sooner or later, 

 in the whole swarm taking wing. During the first movements no definite direction 

 of flight is apparent, but since each individual tries to follow its nearest fellow, a 

 common direction of flight must necessarily result. During the first few days these 

 flights are rather irregular, and swarms do not assume a definite direction, but simply 

 circle above their breeding grounds. If two swarms meet, they mix together, and 

 so the swarm gradually grows larger and larger. The larger the swarms grow, the 

 more regular and the longer become their flights, and at last the time comes when they 

 assume a definite direction, and the insects take leave of their breeding region 

 altogether ; then only few scattered swarms remain where just a day or two before 

 locusts were numberless. In fact this emigration from the breeding regions is often 

 so complete that only single individuals are left behind, and those prove to be nearly 

 all parasitised by the larvae of Sarcophagid flies, or by red mites {Tromhidhim) . 



What is the cause of this emigration ? The generally accepted theory is that 

 locusts migrate from want of food. I have already proved that this is not the case 

 in the larval swarms, and as for flyers, the very idea of locusts being compelled to 

 emigrate from breeding region? by the lack of food could never occur to anyone who 

 has seen these vasts areas overgrown with luxurious vegetation of a kind most 

 acceptable to locusts. In fact, swarms always leave behind them immense feeding 

 grounds and emigrate sometimes to almost vegetationless deserts, which, as a rule, 

 adjoin the permanent breeding areas of migratoria. 



* The habits of the larval swarms of migratoria are well known to those concerned with 

 locust control in Russia, and as the insects are now almost exclusively destroyed by spraying 

 with arsenical insecticides, which are of use only when the actual food of the larvae is poisoned 

 before feeding, no spraying is done during day, when swarms are in movement. The daily work 

 is usually divided into two periods, and spraying is done in the early morning, before the larvae 

 begin to go down from the plants, and in the evening ; the evening work begins before the 

 swarms stop for the night, since it is always easy to reckon where a certain swarm will stop, 

 and the spraying is continued till it is quite dark'; this evening spraying is the more effective. 

 In cool weather, when swarms do not move, work goes on all da}- continuously. 



