Marvels of Pond-Life. 71 



the appearance of the materials employed. All large 

 particles are rejected, and only those retained which 

 will form a homogeneous pulp with the viscid secretion ; 

 and when the process is complete the head of the crea- 

 ture is bent down, and the pellet deposited in its appro- 

 priate place. Each pellet appears originally to possess 

 a more or less conical figure, but when they are 

 squeezed together to make a compact wall they all tend 

 to a hexagonal form, by which they are able to touch 

 at all points, and any holes or interstices are avoided. 



According to Professor Williamson the young 

 Melicerta commences her house by secreting " a thin 

 hyaline cylinder/' and the first row of pellets are 

 deposited, not at the base as would be expected, but in 

 a ring about the middle of the tube. " At first new 

 additions are made to both extremities of the enlarging 

 ring; but the jerking constrictions of the animal at 

 length force the caudal end of the cylinder down upon 

 the leaf, to which it becomes securely cemented by the 

 same viscous secretion as causes the little spheres to 

 cohere." 



Round the margins of the lobes or expansions may 

 be seen delicate threads towards which others radiate ; 

 these are thought by Mr. Gosse to be portions of a 

 nervous system, and two calcars or feelers serve as 

 organs of relation. The young Melicertas are likewise 

 furnished with a pair of eyes, which are probably 

 rudimentary, and disappear as they grow up. 



The Melicerta tubes, being large enough to be visi- 

 ble to the naked eye, are easily crushed in the live- 

 box, and to avoid this, they are conveniently viewed in a 



