65 



The small crab-holes, which are only uncovered at ebb-tide, 

 yielded the larvae as well as the large holes. These larvae live 

 in flooded areas, and it is practically certain that they bury them- 

 selves in the mud, and if this be exposed long- enough to dry out, 

 they pupate and a few days later the Hies emerge. Laboratory 

 observations show that the larval stage may be very prolonged, 

 whereas the pui)al stage is very short, which accounts for the few 

 pupae found. The process of oviposition is difficult, if not im- 

 possible, to observe, but the author thinks that the eggs are laid 

 on the exposed and partially dry mud. He at first failed to find 

 any other midg'e larvae, but at length succeeded in discovering- 

 two entirely marine species, one a Forcipomyia, and the other a 

 Ceratopogon, under the mud and the algae covering it; pupae 

 were easily found under the algae covering- the aerial roots of the 

 mangrove. Other larvae, closely resembling- ('. reticulatus, were 

 found which had a tendency to bury themselves in the mud; one 

 small, possibly C macullthorax, AVilliston, and another larger, 

 C. insignis, sp. n. These and two other species constitute the 

 most numerous and most troublesome of the midges ; the pupae 

 were found in larger numbers by skimming oft' the top of t]i(> mud 

 in the swamp with a shovel. 



The author next describes methods for capturing the Hies in 

 large numbers. The use of the net was found to give good 

 results for the fresh-water forms. The flies taken on man and 

 the larger animals were mostly females. The search for Hies, 

 and especially fully fed females, in houses, stables, &c., did 

 not yield any great results, though they were found in variable 

 numbers on the walls of dark stables in the morning and often 

 when their presence was least to be expected; both sexes were 

 caught on windows, though not often, and the webs of spiders 

 were frequently found to be full of them. No large swarms were 

 ever observed, as is the case with other Chironomids. Lantern 

 traps gave by far the best results. A candle in a spring socket 

 with a glass shade and a metal trough round the outside filled 

 with a mixture of water, alcohol, and glycerine proved a very 

 efficient apparatus for capture. A 50 c.p. electric light, attached 

 to a post 9 metres above ground in the open, yielded good results 

 and a number of new species of Cevdtojxjgoii and Palpo)ni/in were 

 taken by this means. 



The general conclusion from these various methods of observa- 

 tion was that the midges appear a day or two before the full and 

 new moon, are plentiful for a few days and then disappear, often 

 more or less suddenly. A high tide may coincide with their 

 appearance in large numbers, but the beginning of their increase 

 is rather connected with the neap tides. In the Bay of Eio de 

 Janeiro the tides are not specially high and their effect is largely 

 disturbed by meteorological conditions, which react on the 

 " maruim " also and produce irregularities in their appearance; 

 the rule given above is therefore of only very general application. 

 The author says that he has observed no periodicity in the species 

 infesting damp woods inland; they appear to be always plentiful 

 and always aggressive, though drought reduces their numbers, 

 and they do not seem to move far from their breeding grounds. 



