Echinoderm Larvce -with Two Water- Vascular Systems. 343 



recorded in the present paper. For we have learned that when the right 

 hydrocoele is small it may develop lobes, but may have no power to cause the 

 formation of an amniotic invagination (fig. 11), and t)herefore no spines are 

 formed, and the wall of the coelom is also unaffected and no dental pockets 

 are developed. We can, thsrefore, have a right hydrocoele without a right 

 amniotic invagination and right dental sacs, but we cannot have these latter 

 structures without a right hydrocoele. 



I possess one larva which has a right hydrocoele developing long lobes and 

 a vestigial amniotic invagination which is entirely devoid of relation to these 

 lobes. 



The fact that a right hydrocoele can co-exist with one or more pedicellariae 

 raises some interesting questions. These pedicellariae are situated externally 

 to the loop of the ciliated band within which the amniotic invagination is 

 formed, so that when they are not formed, this non-appearance cannot be due 

 to the right hydrocoele taking up the space which is normally occupied by 

 them. It seems far more probable that just as a hormone emanating from 

 the hydrocoele causes the formation of a right amniotic cavity, so another 

 hormone derived from the same source tends to inhibit the formation of 

 pedicellariae. When pedicellariae are nevertheless developed on the same side 

 of the larva, this seems to be dne to the arrested growth of the hydrocoele 

 bud, so that when it really does begin to grow, the normal development of 

 the right side has proceeded to such lengths as to be irrevocable, /o?- in every 

 larva which has pedicellarice on both sides a hydrocoele is cdso present on both 

 sides altliough one of them is ihsucdly vestigial. 



In 1917, owing to a temporary arrest of the growth of the Nitzschia in the 

 plunger jars, the development of the right hydrocoele stood still for about 

 10 days and in the meantime the ominous knobs which are the first indications 

 of the formation ®f the pedicellariae made their appearance on the right side 

 of the larvae, and I feared that all chance of the formation of a well-developed 

 right hydrocele was lost. But to my intense delight this fear proved 

 groundless, for when a more abundant food supply was available the right 

 hydrocoele resumed its growth in spite of the fact that by that time one or 

 even two pedicellariae had been formed. But the occurrence of larvae in 

 which there were present a hydrocoele and two pedicellariae on each side 

 enables us to go further in our analysis. Such a larva can be explained only 

 if we assume that the influences emanating from a hydrocoele not only tend 

 to inhibit the formation of pedicellarite on the same side but to determine 

 their formation on the opposite side of the larva. If we assume that in these 

 larvas the growth of both hydrocoeles had been arrested at an early stage, but 

 after the stage at which the stimulus to form pedicellariae on the opposite side 



