A REVISION OF THE GENUS LOCUSTA, L. 157 



of typical swarm-form voetgangers. The only conclusion I could come to was that 

 these stray forms have been picked up by the swarm. I have also found grass- 

 hopper-phase adults in a swarm of swarm-phase flyers, and have seen a very small 

 green-marked male in copulation with a large typical female. 



" These single forms of L. pardalina can readily be distinguished in the field 

 from our other veld grasshoppers (1) by the fact that they have milky-white glistening 

 underwings, and (2) by their peculiar manner of flight. They almost always soar 

 upwards and then dip and swerve before settling down. 



" From the swarm-phase the grasshopper-phase of L. pardalina differs chiefly 

 in size and in colour. As regards colour, the single forms are remarkable for their 

 great variability, and it would be quite an undertaking to describe in detail all the 

 shades of colour they exhibit. Usually there is a striking protective resemblance. 

 Where there is plenty of green grass both voetgangers and flyers may be almost entirely 

 green, or at least partly green. Where the veld is only sparsely covered with grass 

 and bushes, they resemble the colour of the soil more or less. In parts of the Karroo, 

 notably Beaufort West and Prince Albert, there are patches of gravelly soil varying 

 from slaty-blue to almost black. In 1917 I was greatly surprised to find a very 

 striking tendency amongst the scattered veotgangers to vary in colour from place 

 to place more or less in accordance with the colour of the soil. When the progenv 

 of these scattered locusts appeared on the same farms in swarms six months later 

 there was no trace of such a protective resemblance ; they all wore the King's 

 regulation swarm uniform ! 



" As far as size goes, the single-living forms are generally considerably smaller than 

 the swarm forms. This is especially true of the males. Some of the males are so 

 small that one can scarcely believe that they belong to the same species as the 

 swarm males. 



" When it occurs in large swarms L. pardalina scatters far and wide over the 

 central plateau of South Africa, but its natural home is in the semi-arid parts of the 

 country. It does not like the eastern Orange Free State, for instance, where there 

 is a fairly good rainfall and a rather dense growth of grass. Its favourable breeding 

 grounds are districts like those in the south-western corner of the Free State where 

 the rainfall is slight and the veld consists of short grass mixed with short Karroo 

 bush {Pentzia). In looking for scattered brown locusts I have got the habit of going 

 to spots in the veld where there is an outcrop of white limestone in the red sand." 



These valuable observations of Mr. Faure's leave no doubt that L. pardalina 

 has, like L. migratoria, two different phases, which differ in morphology and coloia- 

 tion, but more profoundly in the biology. Especially striking is it that there is a 

 sort of parallelism in the variation from the swarming to the solitary phase in both 

 these species, as will be presently evident. My study of extensive series of b6th 

 phases of pardalina sent by Mr. Faure, with a careful designation of the conditions 

 under which each particular lot was collected (i.e., whether from swarms or singly), 

 enables me to state the following differences between them. 



The difference in the shape of the pronotum is well marked, though less striking 

 than that between migratoria (or migratorioides) and danica. The pronotum of the 

 swarming phase of pardalina (fig. 5, A & B) is more constricted before the middle, 

 with the fore margin feebly prominent, the hind angle distinctly rounded, and the 

 median keel slightly lower, and deeper cut by transverse sulci than in the solitary 

 phase (fig. 5, D & E) ; but it is hardly possible to express these differences in figures, 

 as I have done for the phases of migratoria, in which they are far more pronounced. 



It is possible, however, to apply the method of proportions to another character — 

 that of the relative lengths of the elytra and hind femora. As in migratoria, the 

 elytra of the swarming phase of pardalina are relatively longer and the hind femora 



