THE LARVAL CASE OF PLATYPHYLAX. 233 



is, in general, a regular arrangement ; the posterior, oldest, part 

 of the case is made entirely of small grains, the anterior, newest, 

 part of the shell of much larger and more irregular grains ; 

 between the two ends there is a gradual increase in the size of 

 the sand grains used. As the larvae increase in size the shell is 

 made wider and larger grains are used in its construction ; after 

 building is completed, or nearly so, there are often two or three 

 very large pieces placed near the anterior end. When the larvae 

 begin to prepare for pupation nearly all the pieces added are of 

 greater size than any of the others, and, finally, when the case 

 is closed, the anterior fourth or third is nearly covered with these 

 large grains, which are so marked and constant that by their 

 presence one can easily tell a larval from a pupal case. Often 

 just before pupation a larva would close the posterior end of its 

 case by the addition of a single large sand grain over the open- 

 ing at this end. After closing the shell for pupation a small 

 opening was always left at the anterior end. This was found 

 between two of the large sand grains, the space between which 

 was filled with a layer of the secretion from the spinning glands, 

 and in this the opening was found. 



Two different kinds of plant seeds were sometimes used in the 

 construction of the cases and these, with the rare use of pieces 

 of snail shells, composed with the sand grains, the materials of 

 construction. In many instances when the seeds had been used 

 in building a case but one or two of them were used ; cases 

 were, however, often found in which were a dozen or more of 

 the seeds and often the two kinds were found in the same case. 

 We had several larvae with these cases living for a few days in 

 the laboratory where the water remained warm and this some- 

 times resulted in the germination of the seeds. When this hap- 

 pened the larvae presented a curious appearance, their cases 

 having a few young plants, one quarter to one half an inch in 

 length, growing on them. 



The experiments we made were upon the construction of new 

 cases by larvae from which the original cases had been removed, 

 and upon the repair of cases after certain parts had been cut out. 

 To facilitate the work and make it possible for us to tell exactly 

 how much new shell had been added, we placed in our aquaria, 



