72 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 



Reichensperger, are very defective, aU the arms apparently being in process of re- 

 generation. The color in alcohol is whitish, with the cirri and proximal portion of the 

 arms somewhat red brown. 



Of the two specunens from the Western Channel of Smigi Barki one is much 

 broken. The arm length is about 40 mm. The cirri are about XXV, 19-20. 



The two large specimens from off Lola have an arm length of 50-60 mm., and the 

 cirri are about XX, 19. Pi has about 9-12 segments. Pj has about 10 segments. 

 The color is dark red. 



The smallest specmien secured by Doctor Merton is that from off Palu Bambu, 

 which has an aim length of 35 mm., and the cirri XVI, 18-20. 



Remarks. — Dr. Torsten Gislen distinguished as form scabra the specimens from 

 Mjoberg's station 4. All these specimens present segments of a more or less rough 

 appearance, and the arms feel scabrous. The median dorsal prominences on the 

 proximal segments are less developed than usual. The basal portions of the arms 

 have the lateral processes more perfected, so that they appear more strongly "wall- 

 sided" than is usually the case. 



Dr. Gislen said that the conclusion to be ch-awn from his study of the specimens 

 collected by Dr. Mjoberg is that the number of cirri and cirrus segments, the length 

 of the proximal pinnules, and the number of the segments is in these pinnules are all 

 features with rather great variability. It is fairly certain that they vary at different 

 ages, and that the figures of variabihty would have been still larger if the available 

 specimens had been more unequal in size. 



Gislen said that the number of the cirri varies between XIX and XXXV. As 

 to their distribution on the centrodorsal, we can distinguish more or less distinctly 

 two categories — an inner dorsal row about the dorsal pole, in which the cirri have as 

 a rule a smaller nimiber of segments, and an outer row about the periphery of the 

 centrodorsal in which the cirri have a greater number of segments. Gislen remarked 

 that this is probably a usual phenomenon, evidently correlated with the fact that the 

 animal is still growing, and also that new cirri are formed at the radials on the border 

 of the centrodorsal. 



Gislen remarked that the yoimger cirri are therefore to be found in the peripheral 

 row, and if it is true that the number of cirrus segments increases during the growth 

 of the animal it is obvious that the older cirri, in the row about the dorsal pole, 

 ought to have fewer, and the yoimger, in the peripheral row, ought to have more 

 segments. 



Though the proximal pinnules vary both in length and in the number of their 

 component segments, the relative length of Pi, Pj, and P3 is nevertheless imiform. 

 The number of segments in Pi varies between 13 and 20, and the length of this pinnule 

 between 10 and 15 mm. P2 is composed of from 13 to 19 segments, and is from 9 to 

 14 mm. long. P3 is shorter and usually has fewer segments. In the pinnules follow- 

 ing the length either does not increase at all, or increases only very slowly, but the 

 number of the component segments through the shortening of the segments rises 

 rapidly to the mmiber (20-25) in the distal pinnules. Gislen noted that the appear- 

 ance and proportions of corresponding pinnule segments are rather closely similar in 

 the specimens described. But there are distinct, though small, distal prominences 

 on the segments of Pb in the eighth specimen described from station 10. Gisl6n 



