308 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 125 



The larval stage lasts i to 9 weeks, depending on the species and the 

 environmental conditions. Thus, the larval stage of Gigantodax aqua- 

 marensis, found breeding in streams with pH about 5, lasted 9 weeks. 

 The larvae of all Guatemalan species whose development has been 

 observed molt six times. In the larval habitat the last instar spins a 

 silken cocoon in which pupation takes place. The form of the cocoon 

 is commonly used to distinguish the species. The pupal stage varies 

 from 2 to 10 days, depending primarily on the temperature of the 

 water. Upon emergence, the adult rises to the surface of the water 

 and takes to flight, soon after which mating is usually accomplished. 



Oviposition usually takes place at the water's surface, or just below 

 it, on some convenient substratum such as bare rock, emergent vege- 

 tation, or debris ; rarely, the eggs are deposited as much as 3 to 6 

 inches beneath the surface of the water. Some species lay the eggs 

 singly, but usually they are deposited in groups that are covered by 

 a viscous matrix. A single female may lay as many as 500 eggs. The 

 young larvae hatch from the eggs in 3 to 20 days. The specific ovi- 

 position habits of the principal anthropophilic species in Guatemala 

 will be discussed along with their developmental cycle. 



METHODS USED FOR STUDYING LIFE HISTORIES 



In carrying out the life-history studies, the results of which are 

 presented below, two techniques were used. The first entailed observa- 

 tion of the species in their natural habitat. The known breeding places 

 of the species to be studied were visited regularly until ovipositing 

 females were found. Detailed data were kept and daily checks were 

 made to observe hatching and the development and habits of the larvae 

 and pupae. The second technique was to observe the developmental 

 stages on the laboratory grounds. A small channel was cut through 

 the sloping lawn of the laboratory and enough water was diverted 

 from a large stream to create a small rivulet similar to the natural 

 breeding places. The same emergent and cover vegetation that is 

 found in the streams in which the species breed, was planted in and 

 along the rivulet. When females were observed ovipositing in the field, 

 the tgg masses were collected and immediately translocated in the 

 laboratory canal. A series of long, raised, shedlike covers were con- 

 structed to protect the canal during heavy rains, thus preventing large 

 numbers of larvae from being washed downstream. With this system, 

 observations could be made on a single species, eliminating the possi- 

 bility of confusing the larvae and pupae of more than one species or 

 of more than one ^gg mass. 



