664 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 



In variety 4 from Ubay {timorensis) the combs are found on the pinnules of the 

 division series, Pi and Pa, and the processes forming them gradually cross over and 

 come to rise from the ventral (nearest the arm) surface of the calcareous segments 

 instead of from their outer margins. 



According to Carpenter the combs consist of from 10 to 12 teeth on a pinnule on 

 the IIBr series, this number decreasing to 6-8 teeth on a brachial pinnule. 



The forms described by Carpenter in 1888 and herein assigned to timorensis or 

 to parvicirra have combs as follows : 



In elongata the first 6 or 8 pinnules have a slight terminal comb which occurs at 

 intervals to far out on the arm. 



In simplex the first 4 pinnules on each side have a small terminal comb which is 

 found at intervals until near the end of the arms. 



In rotalaria (from Zamboanga in 18 meters) the terminal comb is rather small and 

 does not extend beyond the seventh brachial. 



In quadrata there is a terminal comb on the pinnules of the first 9 brachials, and 

 thence irregidarly until the twentieth. 



In his final description of parvicirra (1888) Carpenter said that the terminal 

 comb varies very much is size, being sometimes quite small and inconspicuous. It 

 may not extend beyond the pinnide of the fourth brachial, or it may occur on all the 

 pinnules to the twelfth, and at intervals to the twenty-fifth, occasionally even to 

 near the end of the arm. 



In valida (timorensis) as described by Carpenter the terminal comb is small and 

 much obscured by perisome. 



In littoralis {timorensis) the proximal pinnules have a well-defined comb which 

 disappears by the fifth or sLxth brachial. 



Carpenter figured 3 variations of the comb in as many different individuals 

 from Zamboanga. In one, as he said, the comb is so small that it might easily escape 

 notice. It seems from the figure to be composed of about 4 teeth. In the next it 

 is larger, with 5 well-developed teeth. In the third it is still larger, with 8 still more 

 developed teeth. 



Dr. H. L. Clark wrote (1921) that he is by no means sure that parvicirra and 

 timorensis (annulata) ought to be regarded as distinct, but, on the other hand, it 

 seemed to him unwise to include them both imder a single name. He noted that many 

 specimens of timorensis have only from 21 to 29 arms and often have cirri, sometimes 

 XVIII or XIX, so that it is only by a careful weighing of all the characters that the 

 2 species can be distinguished. Doctor Clark believed that, aside from the matter 

 of color, parvicirra is distinguishable from timorensis by its lighter and more fragile 

 structure, by the presence of some cirri, and by the relatively small number of arms. 

 He concluded that perhaps further observations on living specimens may reveal a 

 more natural line of division than is at present known. 



The habitus of parvicirra is much more variable than that of timorensis; while 

 the individuals are always small with relatively few arms, they are often robust. 

 The cirri as a rule are more numerous and better developed than in timorensis, but 

 they are often few, and occasionally wholly lacking. 



Gisl^n in 1922 divided parvicirra into 3 "subspecies," a comasteripinna, /3 

 comanthipinna, and 7 vanipinna. He did not formally diagnose these forms, but in 



