SECTION I. — INTRODUCTION. - 17 



This was found to be correct, but once more tlio few enemies found 

 A\ ere the same as those already known in Trinidad. {See p. 43) 



In December 1916 I visited the Island of Grenada about 90 miles 

 North of Trinidad, to see some fields damaged by froghoppers, and 

 found the Trinidad insect, T. saccharina, sevei'ely damaging canes in 

 one or two small areas in the uplands of the island. A. report on this 

 outbreak has already appeared (Williams, 1918 A.) In a very hastv 

 visit no insect parasites were seen and one parasitic fungus, the Green 

 Muscardine, already known in Trinidad, was found to be present, but 

 very scarce {see also p. 44). 



In January 1917 I left Trinidad for Panama, visiting on the wa\- 

 La Guayra, Venezuela, where in a few hours ashore a species of frog- 

 hopper, Tomaspis ijropin-^ua. Walk, was found on grass. This is 

 closely related to, but distinct from the Trinidad insect. 



In the Canal Zone, Panama, I found two froghoppers, Toniaspis 

 lepidior and an undescribed species, both on grass and both related to 

 the Trinidad insect, but as weather conditions were very dry and not 

 favourable to the search for parasites, I left the Canal Zone and went to 

 Bocas-del-Toro, on the Caribbean Coast, just at the border line of 

 Panama and Costa Rica. 



In this district I spent four and a half months and investigated the 

 habits and parasites of ten species of the genus Toniaspis. Four of 

 these were forest inhabiting insects, one fed on the roots of Heliconia, 

 another on the attaching roots of an aroid creeper, and a third on the 

 roots of a forest bush, and as these habits were so different from the 

 Trinidad T. saccharina that it was improbable that their parasites 

 would be of any practical value, little attention was paid to them. 

 The remaining six species were grass feeders and two of them were 

 closely related to T. saccharina. A close study was made of these with 

 the result that once again nearly all the known Trinidad parasites were 

 found, including the two specific ones, the Syrphid Fly {SaJpingogaster 

 nigra)\vh.iQ,h. attacks the nymph, and the Vermilion Egg -parasite {Oligosita 

 girauUi). Once more the wide distribution of the parasites was 

 emphasised, as opposed to that of the host insects, few of which seem 

 to range over more than a few hundred miles. After three months 

 work in this locality, a new egnj-pn-asito was discovered, a Mymarid. 

 Anagrus sp. Three specimens of this emerged from some froghoiiper 

 eggs which had been submitted a month before to a miscellaneous 

 collection of small Hymenoptera, obtained, by means of parasite boxes, 

 from grass m which froghoppers had been abundant. 



Unfortunately these three specimens emerged on one day during my 

 absence on a short expedition to find new froghopper localities, and 

 were all dead when I returned to my headquarters on the following day. 

 The loss of these specimens was entirely due to an unfortunate 

 attempt to carry out single-handed an investigation in which at the 

 very least two people should have been engaged. 



Six weeks more were spent in a search for further specimens in the 

 locality which had produced the first three but without avail, and an 

 examination of some thousands of small parasitic Hymenoptera already 

 collected, revealed only two more specimens of this species of Anagrus. 

 Not only, apparently, is it very rare in this locality, but it is so similar 

 to anoUier extremely abundant species of the same genus that it wmild 



