4. Pub. Puget Sound Biol. Sta. Vol. 2, No. 32 



partially into view, its development evidently interfered with by the pre- 

 oral lobe. The oral aspect of the starfish therefore shows a definitely 

 developed oral disk and a eircumoral ring, but no mouth as yet. It also 

 shows external evidence of the ring canal, from which the radial canals 

 extend outward in each ray between the 2 pairs of tube-feet fundaments. 

 Five apices of the hexagonal starfish are now distinctly visible. The 

 further development of the 6th ray is dependent upon the retrogression of 

 the preoral lobe (Fig. 13). This organ has by this time assumed a very 

 variable shape, the larval arms being in various stages of resorption into 

 the main body of the preoral lobe. New tube-feet are added so that gen- 

 erally when the ith pair of tube-feet fundaments appear in the first 5 rays 

 the second pair may have ajjpeared in the 6th ray. There is, of course, 

 considerable variation in different individuals of the same brood and still 

 greater variation in different broods. The ring and radial canals becom-^ 

 more prominent, and indications of a mouth can be made out. 



The 3 arms of the preoral lobe gradually shrink and completely merge 

 into the main portion of this organ, which assumes a form as shown in 

 figures 14" and 15. The fixing disk becomes constricted into a terminal 

 knob, which for a time still serves as an organ for attachment. It finally 

 becomes so nearly separated from the rest of the preoral lobe as to suggest 

 its final discard. It is possible that it may persist attached until the pre- 

 oral lobe is all gone, but the writer has looked in vain for positive evidence 

 of its final resorption. As the preoral lobe shrinks the interval between 

 the first and 6th rays more and more approximates that between any of 

 the other rays, giving more and more freedom for the development of 

 the 6th ray, which appears to be making up for lost time and to be gaining 

 perceptibly on the other rays. However, even after the preoral lobe has 

 entirely disappeared there may be detected for a short time a slight but 

 rapidly waning inferiority of the 6th ray. The position of the madrepore 

 and the anus cannot be made out by superficial scrutiny alone, even after 

 the metamorphosis is complete; but in stained specimens these organs can 

 be located as shown in figure 14. The Madrepore therefore lies in the 

 ■nterray between rays 1 and 2, which agrees exactly with Gemmill's (1912) 

 observations on Solaster. Before the larval organs of attachment entirely 

 disappear the tube-feet become functional. Their capacities for extension 

 and retraction and for coordinate movement are very striking. The mouth 

 is also open at this stage as proved by the protrusibility of the stomach. 

 Terminal and aboral spines develop in regular fashion, later to be fol- 

 lowed by ambulacral spines also. All these arise from previously formed 

 skeletal plates that may be readily demonstrated in cleared specimens. 

 They may even be seen in the latest stages of metamorphosis in the living 

 .mimal when the disappearance of the yolk supply renders the individual 

 whiter in color and more translucent. 



