2 JAMES WATERSTON. 



during the summer, e.g., Tringa (Totanus) ochropus, T. hypoleucus and Aegialitis 

 minor were constantly observed feeding in shallow pools on mud flats or in woods 

 near the Vardar, and the absence of larvae in these situations was remarked. That 

 larvae had been previously present could not however be ascertained. 



Various small species of fish and the fry of larger forms control the larvae to some 

 extent. Yet even when fish were present, very little cover seemed to afford the 

 larvae protection ; for example, both fish and larvae were found in pools above 

 Likovan. In other cases, especially in clear gravelly or rocky pools, the fish-control 

 appeared to be effective. In the Gumus Dere in September I found a few larvae 

 and no fish in some of the upper parts, but as one went towards the Struma the fish 

 increased and the larvae disappeared. In Ardzan Lake the clear reed pools are 

 patrolled by shoals of fry and I found no larvae there ; on the other hand small 

 fish were found on the lakeward side of the barrier of broken reeds, etc., behind 

 which larvae of Anopheles sinensis abounded. Several species of small fish occur, 

 but only one, Gobius rhodopterus, Gunth.,* has as yet been identified. 



Under laboratory conditions many insect larvae collected with those of mosquitos 

 were predatory on the latter, e.g., various Odonata, Ephemerids, Chrysops, etc. 

 Nepa cinerea and Notonecta glauca also took their toll, but the evidence was 

 confusing as to what takes place under natural conditions. In some pools, for 

 instance, Notonecta swarmed and little else living could be found ; in other cases 

 larvae occurred with the bugs. A good deal depended apparently on the size of the 

 sheet of water and whether the sides were clear or fringed with weed. One small 

 pond in the Struma Valley, about 40 yds. long by 6-7 yds. wide, produced a steady 

 supply of larvae and had ultimately to be oiled. Yet the water teemed with dragon- 

 fly, Ephemerid and Coleopterous larvae, and after oiling, the following large water- 

 beetles were secured : — Cybister lateralimarginatus, de G., 8 $, 16 $ ; C. tripunctatus, 

 Oliv., 3 <J ; C. tripunctatus var. gotschi, Chd., 1 3; Dytiscus marginalis, L., 1 $ ; 

 D. dimidiatus, Berg., 1 $; Colymbetes fuscus, L., 1 £ ; and Hydrophilus piceus, L., 

 1 ?. I have quoted the numbers, which are noteworthy in so small a sheet of water. 

 Not all the specimens killed by the oil were secured. 



Adult dragonflies also account for many mosquitos, although I have seen only 

 Culicini hunted. About 15 species of dragonflies were secured by only intermittent 

 collecting, and of these the most active were species of Anax, Orthetrum, and 

 Sympetrum. 



Towards the end of the season larvae became badly overgrown with colonies 

 of Vorticella. These however did not prevent the completion of development. 

 There was also a tendency for the long branched clypeal, lateral, terminal and 

 palmate hairs to break off. Under these circumstances some examples were unable 

 to grip the water surface sufficiently and appeared to drown. Many other individuals 

 that died showed shortly before this an abundant growth of a fungus allied to 

 Saprolegnia ; but whether the fungus was a primary or secondary factor in this was 

 not determined. 



Throughout the autumn occasional mosquitos with acari were captured. Such 

 cases occurred more frequently towards the end of the season. 



♦Determined by Dr. C. T. Regan, F.R.S. 





