588 Cuin Pine 
The writer observed the following interesting phenomenon relative 
to respiration. The water in the aquarium was thick with microscopic 
organisms and was greenish in color. Each of the several larvae crawling 
at the bottom had a bright silvery air globule attached to the caudal end. 
The size of that globule was as large as the head of the larva. Soon the 
globule became larger, and the larva floated up with it. As the larva 
touched the water surface with its caudal end, the silver globule suddenly 
burst and the larva immediately sank down to the bottom. Sometimes 
it took about ten or even thirty seconds for the larva to break the globule 
after it reached the surface. While the larva was in its course toward 
the surface, it lay straight, curled obliquely, or wriggled and clasped 
another larva with the last pair of prolegs. But whatever the position 
or movement the caudal end was always pointing upward. The air 
globule began to form when the larva was sinking down halfway or some- 
times very near the bottom, and its size continued to increase there- 
after. After the larva had crawled along at the bottom for about 5 or 6 
centimeters, the globule gained its full size and the larva was ready to 
float up again (fig. 73). This rising and sinking of the larvae kept them 
restless and gave the aquarium a most lively appearance. It is be- 
lieved that such action as this takes place only when the conditions in 
the aquarium are getting abnormal and unfavorable for respiration. 
Preference for stagnant and shallow water 
The conditions at the pools and overflowed areas are, throughout the 
season, admirably favorable to the life of the larva. It prefers stagnant 
water, because it wriggles and suspends itself under the surface, and, 
because of the absence of specially adapted apparatus, it is unable to 
gain foothold or to pursue its course in rapid streams. ‘The larva pre- 
fers shallow water, because it is easy to reach the surface, where respi- 
ration takes place, and because the floating scums offer food supplies. 
The pools have no outlets, and the water, which is accumulated from 
rains in the summer season, remains permanently still and has never 
exceeded 40 centimeters in depth, so that the development of the larval 
life here is much favored. The presence of larvae in great numbers in- 
dicates that these pools serve as an excellent habitat. At different times 
of the season the salinity and density of the water vary greatly. Some 
of the pools have a sudden decrease of larvae owing to the over-increase 
