Semi-Centennial Volume. 71 



"Upon picking these sticks out of the water shortly afterwards, I al- 

 most invariably found that a larjre portion of the larva? had disappeared. 

 On proinp back to the place on the next dav, I found very few larvae 

 attached, althoug:h the pup.ne still remained in numbers. The larva? seem 

 to realize when they are in dang^er, and as Foon as an infested lo^ which 

 has been taken out of the water is restored to its former position, they 

 release themselves from the lo^ and float down stream to other quarters. 

 All of the larvas do not immediately forsake the lop:s thus replaced in 

 the water, the process of desertion beine: gradual, but a p^reat propor- 

 tion of the larva; seem to leave their place of attachment. I observed 

 this characteristic frequently after this time. 



"Miall, in describing one of the British species which attach to sub- 

 merged leaves, says that this feature of releasing themselves is a normal 

 habit, but that the larvae generally remain attached to their supports 

 by the silken threads which they spin, and shortly after being dis- 

 turbed, draw themselves back to their original locations. I do not think 

 this action is often true of Sinndivm pecnarnm. This feature of their 

 behavior made observations much more difficult than I had expected. 



"In order to determine whether the insects bred in still water, I 

 examined numerous twigs and branches in the backwater of the bayou 

 and in several standing water holes and other places between the bayou 

 and the back land, and also many apparent ideal places of attachment in 

 the stagnant water of the back land, but never found the slightest indica- 

 tion of a gnat stage. In all, I examined several hundred possible places 

 of attachment. 



"To further test this point, I collected numbers of larvae from the 

 bayou, leaving them attached to their original supports, and placed them 

 in fresh bayou water held in quart and half-gallon jars with the top 

 left uncovered. Many hundreds of larvae were given an opportunit\' to 

 live in this way, but I never succeeded in keeping any alive for moie than 

 twenty hours, the length of life when thus confined varying from six or 

 eight on to nineteen and twenty hours, though few larvae lived the latter 

 time. I also tried confining larvae singly in small phials with bayou 

 water, unattached to an^ thing, and found that they lived in this way 

 about as well as in the iars. They at first attached to the sides of the 

 phials, and generally climbed up later to the top of the water, where 

 by means of the posterior sucking apparatus they hung head downward, 

 being suspended from the surface film of the water. Upon dying, being 

 heavier than water, all Sitnuliinn larvae generally sink to the bottom of 

 the vessel. In the jar experiments, the greater number of the dead 

 larvse were always found upon the bottom of the vessels. Larvae hanging 

 to the surface film of water will, when dead, remain floating there until 

 the film is broken. 



"March 17. — Secured a boat and paddled up and down Bayou Pierre, 

 trying to determine just where SimnUum pecnnrum would or would not 

 breed. After examining a great number of submerged and floating twigs 

 in the swift current, and partly out of it, and also those plants growing 

 near the banks, I came to the following conclusions: The gnats are 

 seldom found en plants fringing the .'■hore or out of the reach of the 

 swift current. I never found a larva on any plants in such situations. 

 However, in nearly every case where a twig in the full current was 

 pulled up, numerous larvae were found attached. The number of larvae 

 seemed to decrease in direct proportion to the strength of the current. 

 Wherever a stick was jammed in between 2 logs through which the water 

 was compelled to flow with a great deal more than ordinary swiftness, 

 the larvae were found most abundant, generally holding on by the pos- 

 terior cup and swaying in the water, and creating small currents towards 

 their mouths by rapid swishes of the peculiar many branched filaments 

 of the head. They often, when shifting, .secure themselves with a tiny 

 silken thread. 



