52 G. F. HILL. 



determined, but the period required for the life-cycle, as ascertained later, suggests 

 strongly that the latter was more probable. Further, there is no evidence to suggest 

 that the Tabanids in this or other districts known to me produce two generations in 

 one season. Obviously these remarks do not appl}/ to localities in which there is 

 permanent surface water, such as Locality No. 1. 



In some localities a few individuals of certain species, particularly T. rufinotatus, 

 are to be found throughout the dry season (June to August), but their occurrence can 

 be accounted for by the presence of permanent water (such as Locality No. 1), or at 

 any rate marshland, in the vicinity. 



During June of this year numerous young larvae were found in pools which dried 

 a few weeks later. The question naturally arises whether these larvae perish or 

 whether they are able to burrow into the mud to he dormant through the remainder 

 of the dry season and until conditions again become favourable for their development. 

 Attempts to determine these questions have been unsuccessful, but it may be mentioned 

 that one larva from a late batch of eggs was, when about 9 mm. long, accidentally 

 isolated in the laboratory in a pot of dry sand and without food of any kind from about 

 6th June to 28th August. On the latter date it had the general appearance of adult 

 larvae in the resting stage preceding pupation, and became active immediately it 

 was placed in water containing suitable food and cover. Unfortunately neglect 

 caused its death before any development could be detected. Hine, working in 

 America, found that certain Tabanid larvae after a long resting period again fed 

 before pupating, but nothing of the kind has been observed to take place in the case 

 of any of the species studied here. 



Rearing Larvae in Captivity. Various methods of rearing the young larvae were 

 tried, but as none of them gave satisfactory results it is not intended to describe them 

 in detail. In order to prevent cannibalism young larvae were isolated in small earthen- 

 ware pots with a capacity of about 120 c.c, prepared in various ways, and suppHed 

 with various kinds of food. x\part from the difficulty of keeping the water in these 

 vessels fresh, this method was found to be too cumbersome, and the larvae rarely 

 survived for more than a week or so. Larger vessels, i.e., five-inch flower-pots and 

 small museum jars, were equally unsatisfactory. Kerosene tins cut lengthwise into 

 two equal parts gave better results, and in them a few larvae were reared from 

 the egg to maturity. These tins were prepared by placing a quantity of clean sand 

 at one end and two or three inches of water at the other. Pieces of water-lily 

 leaves, algae and swamp plants carrying small molluscs were placed in the water 

 to afford shelter and food, and these were renewed as often as possible. The water 

 was changed every four or live days by lifting one end of the tin and allowing it 

 to filter through the sand. The tin was then partly filled and again emptied in 

 this way, before being finally replenished with water and food-bearing vegetation. 

 Each tin contained the progeny from one batch of eggs, i.e., 300-600 larvae, of 

 which never more than 1-2 per cent, reached the imago stage. When molluscs 

 were not obtainable mosquito larvae and small earthworms were offered as 

 substitutes, but worms were invariably refused by larvae in all stages of their 

 development. Mosquito larvae were destroyed by the Tabanids when the former 

 were stranded in algae or sand, but otherwise they appeared to have been able to 

 avoid capture. At all times Tabanids of this species appear to feed upon their fellows 

 in preference to molluscs or any other animals, and for this reason it was found best 

 to transfer the survivors, when of nearly full size, to separate pots or dishes, where 

 they completed the development in moist sand, upon which molluscs and portions of 

 lily leaves were placed. The best results were obtained from a batch of 105 larvae 

 which were reared in a concrete trough, measuring 16 inches wide by 20 inches long 

 by 6 inches deep, prepared similarly to the tins just referred to. As the larvae 

 approached maturity the water was gradually reduced until it was confined to a small 

 area at one end of the trough. In this molluscs were placed every week or so, until 



