338 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



by the same curious process of twinning, through which one of the resultants arises 

 much later than the other, that we see illustrated everywhere throughout the 

 crinoid organism. At the time of the extrusion of the radianal from the radial cir- 

 clet a very narrow plate appears separating the two posterior radials. Almost 

 simultaneously four other similar plates appear separating the other radials in the 

 four other interradial areas. These plates are narrowly oblong, their longer sides 

 in contact with and equal in length to the lateral edges of the two radials which they 

 separate. All these interradials rapidly increase in width, and just beyond their 

 distal border two delicate plates appear as flat, more or less rhombic, calcareous 

 films, the smaller just beyond the larger. At this time the five primary postradial 

 series are fairly well developed, possessing numerous brachials beyond the IBr 

 axillary. These two filmy plates increase in size and gradually transform into 

 a IBr series from which two arms are given off; and these in every particular, except 

 size, resemble the IBr series and arms borne by the primary radials. It is not until 

 the animal is fully grown that the five interradial postradial series attain the size 

 of the five primary postradial series, and the five interradials assume all the characters 

 of true radials. 



The interesting Thaumatocrinus renovatus (figs. 115-118, p. 183) is the young 

 of the species later described as Promachocrinus abyssorum (with which it was 

 found associated) just after the resorption of the radianal and the formation of all 

 of the interradials from which the five additional arms are commencing to grow. 

 The posterior interradial arm as seen in the so-called Thaumatocrinus is the first to 

 form, and is consequently larger than the others ; but from the size of this posterior 

 arm and the breadth of the interradials I suspect that smaller arms borne on the 

 other interradials have been lost, as these interradial arms when small are extremely 

 delicate. During growth the posterior interradial arm of Thaumatocrinus becomes 

 reduplicated on all the other interradial plates, and all of the five interradial arms 

 gradually increase to the size of the five primary arms (the extensive plating of the 

 disk at the same time disappearing by resorption) so that the 10-armed Promacho- 

 crinus abyssorum results. 



Anal x in the fossil forms may be reduplicated hi the form of a series of inter- 

 radials, one in each of the interradial areas, and therefore, bearing in mind the 

 greater perfection of the radial symmetry in the recent types, it does not surprise 

 us to see the same thing in the recent comatulids. 



In some thirty 6-rayed specimens which I have studied the supernumerary ray 

 is in all cases but two inserted behind the left posterior that is, between the two 

 posterior radials and receiving its ambulacra from the groove trunk to the left. It 

 is impossible to interpret this otherwise than as the persistence and subsequent 

 development of anal x in types in which the interradials, including anal x, are 

 normally resorbed immediately after formation, exactly as it is developed in Pro- 

 machocrinus and Thaumatocrinus. Additional weight is given this view by the 

 fact that Promachocrinus kerguelensis is very often 6-rayed, the additional ray 

 being in that case inserted behind the left posterior; only anal x has been formed, 

 the other interradials either having been entirely suppressed or having been, as in 



