238 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



descriptions of Lesueur (1813 and 1815) and Savigny (1816) materi- 

 ally advanced knowledge of the genus, and Huxley's studies (1851, 

 1860, a, 1860, b, and 1861) gave a still more adequate conception 

 of its organization as well as of the reproduction. There has been 

 much confusion between P. atlanticum atlanticum and the several 

 other subspecific forms. We will make no attempt to follow through 

 all these descriptions and resolve the confusion. The attempt would 

 be tedious and at best only partly successful. 



In our work on the extensive collections of Pyrosoma, made by 

 the United States Bureau of Fisheries, we have been led to believe 

 that no valid distinctions can be made among Atlantic forms of 

 Pyrosoma, which would show the existence of the two nearly allied 

 subspecies, atlanticum, and giganteum, generally described. In the 

 Pacific Ocean, however, ihere is found in comparative abundance a 

 Pyrosoma which agrees with descriptions of P. atlanticum from the 

 Pacific Ocean. It may occur in the Atlantic Ocean as well, but we 

 do not find it among our 43 colonies from the Atlantic Ocean, 42 of 

 which, collected at 30 stations, belong to the major atlanticum -group, 

 nor are there published descriptions which indicate the presence in 

 the Atlantic Ocean of P. atlanticum atlanticum, as we define it. It 

 seems best to recognize as giganteum the large colonies from the 

 Atlantic Ocean so often described under that name. 



We find, also, in the Atlantic Ocean less specialized forms of 

 Pyrosoma than the typical giganU wm, which appear to be intermediate 

 between the Pacific type, P. atlanticum atlanticum, and the genuine 

 P. atlanticum giganteum. These will occasionally be referred to for 

 convenience as P. atlanticum, group intermedium, pending the study 

 of further collections, which may indicate that some of these are 

 distinct forms which have not reached full development. 



The major species under consideration has been most often spoken 

 of as the typical representative of the genus. It has ordinarily been 

 described as having a conic-cylindrical colony, beset with truncated 

 test processes or spines; the zooids irregularly arranged in the test, 

 and, for the most part, long; with other general data, in no way dis- 

 tinctive. This brief diagnosis, while accurate enough, serves only to 

 set off a whole group of Pyrosomas, containing many diverse varieties. 

 These, while easy to identify if characteristic specimens are at hand, 

 show such intergradation one with another, and conform so closely 

 with the fundamental plan for the group, that we are treating them 

 as varieties or subspecies of a single species which we name P. atlan- 

 ticum. Taking up, then, the several nearly allied subspecies, we 

 treat first the form chosen as the species type, probably the same form 

 Huxley (1851) described as P. atlanticum. The selection of this 

 form as the species type seems natural, since, during the immature 

 stages of the growth of the colony, some of the other subspecies 



