562 



of the permanent breeding grounds of migrator ioides. The gregarious- 

 ness of the migratorioides phase is one of the causes of a rapid increase 

 in the number of individuals and swarms ; in the course of a 

 few generations the size of the swarms reaches the maximum, and 

 this is followed by emigration. In this way the dispersion of the 

 species is carried on alternately by the gradual spreading of the danica 

 phase and by the periodical extensive emigrations of migratorioides. 

 As a result," the species is now distributed all over the Eastern 

 Hemisphere. 



In the Palaearctic region the swarming phase is represented not by 

 migratorioides, but by migratoria, which stands, as regards its morpho- 

 logical characters, between migratorioides and danica. This fact is 

 explained by the impossibility of finding in the Palaearctic region 

 natural conditions exactly like those of the tropical breeding grounds 

 of migratorioides, chiefly in regard to a combination of heat and 

 dampness. The nearest approach to tropical conditions in the 

 Palaearctic region is represented by vast reed-beds of Phragmites in 

 the deltas of the large rivers emptying into the Caspian and Aral 

 Seas and Lake Balkhash, and there the permanent breeding grounds 

 of migratoria are concentrated. Though very peculiar, and in summer 

 recalling the tropics, the climatic conditions of these reed-beds are not 

 tropical, and the effect on the progeny of danica is not the same as 

 in the tropical breeding grounds of migratorioides ; the reverse trans- 

 formation of danica into a swarming phase does not reach the extreme 

 phase of migratorioides, but stops half-way at the migratoria phase. 

 This seems to indicate that the transformation is due primarily to the 

 direct influence of external conditions, its extent being proportional 

 to changes in the latter, but only precise laboratory investigations 

 can clear up this complicated problem. The development of migratoria 

 in the permanent breeding grounds goes on, as is the case with 

 migratorioides, alternately with periodical emigrations, which are due 

 not to the lack of food resources, but to some internal causes, and are 

 followed by transformation into danica. 



The theory of phases suggests the possibility of the control of 

 migratoria by some measures directed not against the insect itself, 

 but against certain natural conditions existing in breeding regions 

 that are the direct cause of the development of the swarming phase. 

 Even a comparatively slight cultivation of breeding grounds leads to 

 the desired changes, rendering the transformation of the hannless 

 danica phase into swarming migratoria impossible. 



A study of Locustana pardalina, Wlk., which is peculiar to South 

 Africa, shows that it also has two distinct phases — the swarming 

 phase, pardalina, and the solitary, harmless phase, to which the name 

 solitaria is here given. The transformation of one phase into the other 

 is proved, but the conditions necessary for it remain unknown, as the 

 breeding grounds of this species have not yet been investigated. The 

 inter-relation between pardalina and solitaria, as regards the morpho- 

 logical characters, coloration of larvae and adults, and behaviour, is 

 very much the same as in danica and migratoria. 



The causes of migration of larvid and adult swarms of migratoria 

 are discussed, and the conclusion is arrived at that lack of food has 

 nothing to do with the movement of larval swarms, which are guided 

 chiefly by two causes — the instinct of gregariousness, and thermo- 

 tropism. The flying swarms, too, are not driven to wander by hunger 

 and they do not even feed much during migration, but exist at the 



