LIFE HISTORY OF HYDROUS ( HYDROPHILUS) TRIANGULARIS. 11 



not enough water plants in the aquarium to work with. After these had been sup- 

 phed she began to spin another case at 1.25 p. m. Assuming a position, back 

 downward, her body almost parallel with the surface of the water and close to it, 

 she held herself in place by means of the floating water plants (Potamogeton and 

 Elodea), and moved her spinneret rapidly to and fro sidewise, at the same time 

 pushing the material backward with her hind legs. After spinning the roof of the 

 case in this manner for a few minutes she turned over without removing her spin- 

 neret. Her body was considerably inclined, with her head some distance below the 

 surface, and she opened her wing cases slightly to supply herself with fresh air at 

 periods varying from 15 to 90 seconds, the shorter period the more common. This 

 mode of taking in air through the lifting of the posterior ends of the elytra is in 

 marked contrast to the method normally used (see p. 29). It required a longer 

 period to make the floor of the case, since it is much more convex than the roof, 



Fig. 2.— Longitudinal section 

 through center of an egg ease, 

 showing a thick layer above the 

 eggs, a floor beneath them, and 

 an open chamber under the 

 floor, in which the newly- 

 hatched larvae gather. 



Fig. 3.— End view of egg case, 

 showing triangular plate with 

 lunate opening beneath it. The 

 floor beneath the eggs inside the 

 case does not reach the side 

 walls, leaving a free passage to 

 the chamber beneath. 



but it was finished by 2 o'clock. The ends of the threads forming the floor were 

 apparently fastened directly at either end to the previously completed roof, leaving 

 a ridge or seam where the two joined. 



When the case was thus far completed, she began egg laying. The roof was 

 covered on the inside with a thick layer of loosely woven silk inclosing large open 

 spaces. From this layer the eggs were suspended with their long diameters vertical, 

 each egg held in place by a thin meshwork of silken threads, which were attached 

 to the thick layer above and to one another and formed a continuous floor below 

 the egg mass (fig. 2) . 



At 2.15 she stopped laying and, after removing her spinneret, swam about 

 for a time, evidently resting. On returning to the case it required several efforts 

 to replace her spinneret in the open end, but once adjusted she began to weave the 

 triangular plate across the upper part of the open end, leaving below it a lunate 

 opening (fig. 3). The plate was considerably thicker than the rest of the case 

 and was fastened on the inside to the layer of loosely woven silk that covered the 



