42 G. F. HILL. 



horses and cows which drank at the lower pools were watched, but no evidence of 

 Tabanidae could be found until 6th January, or two clays after the first heavy fall 

 of rain (106 points) since the precechng March. 



On 6th January a female T. aprepes was observed on a twig about four and a 

 half feet over the water (Plate I), apparently about to oviposit, when she was 

 disturbed by a spider and flew off. 



On 13th January three egg-masses were found on the same twig, all of which 

 appeared to have hatched, but on removal to the laboratory three living larvae were 

 rescued from a spider's web which enveloped one end of one mass. Although there 

 appeared to be no unhatched eggs in the masses, the latter were placed over 

 water, and during the night 110 larvae were produced, some of which were subse- 

 quently reared to the fly stage and identified as T. aprepes. 



From 16th January until 16th April numerous larvae of T. aprepes were found 

 in the algae floating on the surface of the water. Heavy floods swept the holes in 

 the locality during the month of April, after which larvae were not found. 



On 2nd April a female T. aprepes was observed to alight on the twig from which 

 eggs were taken on the 13th January, and the process of 'oviposition was observed, 

 with the aid of a magnifying glass, from commencement to conclusion, when the fly 

 was captured and the eggs removed to the laboratory, where larvae were subsequently 

 reared. 



Of the 39 egg-masses collected here (Locality 1) between 13th January and 9th 

 April, five were found on twigs and six on grass leaves or seed-heads overhanging 

 water in the rock-hole (Plate I), the majority being from 3 to 4| feet above the 

 surface ; three were found on grass leaves a few inches above the water trickling from 

 the rock-hole, and 25 on the terminal shoots of couch-grass overhanging a sloping 

 bank (Plate II), the surface of which was oozing with soakage water and algal growth. 



The adjacent swamp (Locality 2) dried earh- in July 1919, and remained in this 

 condition until 4th January 1920. Towards the end of April 1920 the surface area 

 reached its maximum, there being then about 4| feet of water in the deepest parts. 

 Repeated searches were made for egg-masses on plants overhanging the margin and 

 on lily and other leaves floating on the surface, but none were found. During these 

 searches numerous half-grown to full-grown larvae {T. aprepes) were found clinging 

 to the lower surface of the lily leaves, or to the stems, or hidden in floating masses of 

 algae, from 20 to 30 yards from the margin and in from 3J to 4| feet of water. A 

 few of these larvae were bred to maturity to confirm identifications. 



The small lily-pond (Locality 3, Plate II, fig. 2) dried late in July 1919, and 

 remained so until 4th January 1920. In May of 1919, when the water was two or 

 three feet deep, a careful search was made for Tabanid larvae amongst the reeds 

 and water-lilies, but none were found. 



From 10th June 1919 onwards, the banks of this and adjoining ponds were examined 

 from time to time as they dried, but in none of them were the larvae or pupae of 

 T. aprepes found, although other species were taken. 



From January 9th to April 16th 1920, when the ponds and swamps contained 

 water, egg-masses and larvae were unsuccessfully sought for on many occasions. 

 On the latter date a careful search was made of the vegetation growing near the 

 banks and of the lily leaves in deeper water (4-5 feet) with the result that many 

 larvae of T. aprepes, in all stages of development, were captured. On the same 

 date a fly of this species was observed ovipositing (1.45 p.m.) on the imderside 

 of a seed capsule of a plant growing in eight inches of water and twelve inches from 

 the bank. When about 30 eggs had been laid, a lamp chimney, closed at one end with 

 mosquito netting, was slipped over fly and plant and tightly plugged with wadding. 

 The stem was then cut off about an inch below the wadding plug, leaving the 



