18 DE. J. P. GEMMILL 0?f THE DEVELOPMENT OF 



larval arms, it is of a more vivid orange-colour than the rest of the preoral lobe. The 

 three arms surround the sucker in tripod fashion, but the right one is now rather 

 larger than the left, which at this stage is somewhat compressed in the angle between 

 preoral lobe and body. 



The actual site of fixation must depend more or less on chance. At any rate, in 

 the aquaria I employed, attached specimens were found fairly uniformly over the 

 bottom and sides alike, a slight preference being shown for a position on the sides 

 just under the water-level. In those which settled on a vertical surface there was 

 no strict uniformity as regards which part of the body was uppermost, but in the 

 majority of instances an upper position was occupied by the neck or region of junction 

 between the preoral lobe and the body. 



Once the larvae are attached they keep the same position till movement is possible 

 by means of the tube-feet. If they are detached without injury during the first day 

 or two after fixation, they will in many instances (five out of seven in the experiment) 

 succeed in re-attaching themselves. They are enabled to do this by the muscular 

 action of the arms, as well as by the action of the cilia on the surface of the body, 

 which helps to tumble them over until the sucker finds a suitable locality for attach- 

 ment. Older specimens when detached sink quickly to the bottom, and can fasten 

 themselves anew only if they happen to fall with the sucker undermost; otherwise 

 they are unable to right themselves, and, though capable of survival for a considerable 

 period, they do not undergo a normal metamorphosis. The first few days of fixation 

 by the sucker are marked by the following important external changes : — 



1. Flexion and torsion of the preoral lobe are completed. 



2. Much of the wall of the preoral lobe is gradually incorporated into the middle 



part of the left side of the larva, and assists in forming the oral surface of the 

 Starfish. Subject to this reservation, and from the point of view of external 

 form, the left side of the larva becomes the oral surface of the Starfish. 



3. Similarly the aboral surface of the Starfish may be described as representing the 



right side of the larva; but this statement is also subject to the reservation 

 that in reality the epiderm of the aboral surface is in great part derived from 

 that which covered the posterior end of the bilateral larva. 



4. The series of radial pouches of the hydroccele comes into prominence on the 



oral (left larval) surface. The first-formed pouches acquire lateral branches for 

 the tube-feet. The 8th radial pouch is added. 



5. The aboral rudiments of the arms begin to grow out round the margin. The 



whole aboral surface becomes covered with small papilla containing the 

 developing spines. 



The later period of fixation by the larval sucker is marked by further outgrowth 

 and commencement of functional activity on the part of the tube-feet, complete 



