PART 5 A MONOGRAPH OF THE EXISTING CRINOIDS 543 



P, is 2 mm. long, very suuill and slender, with 9 segments, of whicb the first is 

 short, the second is slightly longer than broad, the third is twice as long as broad or 

 slighth' longer, and the fourth and fifth are the longest, 4 or 5 times as long as broad; 

 on some of the Pi the remaining segments are of about the same length as the fourth 

 and fifth, while on others they are much shorter. P2 is similar, but slightly smaller 

 and shorter. P3 is 3.5 mm. long, with 9 segments, of which tlie fourth and following 

 are much elongated; the segments have a spinous dorsal surface and expanded and 

 overlapping distal ends which are fringed with spines; the third to fifth segments bear 

 a large gonad. P4 is 4 mm. long, with 10 segments, and is similar to P3. P5 is similar 

 to P4. Pe sometimes possesses a small gonad, but usuallj^ is without one. The distal 

 pinnules are very slender, 5 mm. long, with 17 segments, which beyond the third are 

 much elongated with swollen articulations and slightly overlapping and finely spinous 

 distal ends. 



Notes. — February 15, 1903; 400 meters. One of the two specimens with this 

 label is the one described above. In the other the arms are 45 mm. long and there 

 are 20 to 25 cirrus segments. 



January 31, 1903; 380 meters. The example is simOar to the 2 just noticed. 



April 17, 1902; 385 meters. The cirri are 7 mm. long, and are composed of 22 

 segments, of which the longest are nearly twice as long as the width of the pro.ximal 

 (narrower) end ; the spines on the dorsal surface of the IBr series and lower brachials 

 are especially long in the median line, so that the arms are subcarinate. The gonads 

 are much enlarged. 



December 12, 1902; 385 meters. The arms are 25 mm. long; the cirri are 6 mm. 

 long and arc composed of 17 to 19 segments, of which the longest are not quite twice as 

 long as the width of the proximal end. 



The small number of cirrus segments and the elongate form of the more proximal 

 woidd suggest that this form might possibh^ be the inunaturc of Eumorphometra hirsuta; 

 ])ut on the other hand the largest individual, which exhibits all the characters described 

 for the t3"pe, is larger than E. hirsuta, which has an arm length of onty 35 mm., and 

 several are sexually mature with eggs extruded from the ovaries. 



When he was studying the crinoids collected by various Antarctic expeditions 

 Dr. D. Dilwyn John found that at least one of his new species of Eumor'phometra, 

 E. aurora, is viviparous. On hearing this, I reexamined the specimens of E. concinna 

 at hand and discovered, much to my chagrin, that I had overlooked the fact that it 

 too is viviparous, with brood pouches at the side of the ovaries. So I sent him a 

 portion of the arm of a female to compare with his new species. 



He found that the brood pouches lie on the aboral side of the pinnules, nearest 

 the arm. The walls are so thm that the contents of the pouches can be seen through 

 them. The eggs in the ovaries and the embryos m the brood pouches are of different 

 sizes and at various stages of development. The largest eggs are oval and as much as 

 0.2 to 0.25 mm. long. The largest of the brood pouches contained thirteen embryos. 

 Seven were without trace of skeletal plates or ciliated bands; they were spherical or 

 irregularly oval, 0.21 to 0.29 mm. The other six were oval embryos with the beginnings 

 of skeletal plates within them. The smallest of these was 0.34 mm. long with no ciliated 

 bands. There were about fourteen small colunmals and a very large termmal stem 

 plate, and five orals and five basals, but no infrabasals. Three other embryos in good 

 condition were slight!}' larger, 0.37 to 0.40 nmi. long and 0.29 to 0.3C mm. broad, and 



