22 DISCOVERY REPORTS 



There seems little doubt, however, in view of the evidence presented in this paper, that previous 

 records of the ' asperate ' form of S. fusiformis from temperate or tropic seas are of the same species 

 as the one that I have renamed S. aspera, because in every paper in which such serrated specimens 

 of the aggregate form are figured they are shown with a muscle plan in which M. IV-V do not join 

 laterally (for example, Herdman, 1888, PL VI, fig. 6; Apstein, 1906, PI. XXVI, fig. 6; Yount, 1954, 

 fig. \2b). Furthermore, where a serrated solitary form is figured it is shown with M. VIII-IX parallel 

 or nearly so (Herdman, 1888, PI. V, figs. 1-4; Ritter, 1905, fig. 14; Yount, 1954, fig. iia). 



There are, however, in the literature, three notable exceptions. Herdman (1888, PI. VI, fig. 5) 

 figures a serrated specimen of the aggregate form in which M. IV-V join laterally, Apstein (1906, 

 PI. XXVI, fig. 6) shows a serrated solitary form with M. VIII-IX joined medio-dorsally, while 

 Metcalf (1918, fig. 80) figures a different view of what appears to be a similar specimen to that of 

 Apstein. Herdman's and Apstein's figures, however, are of specimens from the Antarctic and so from 

 the data presented in this paper conform not to S. aspera but to S. thompsoni. The figure in Metcalf 's 

 paper is not original, being based on that of Ritter (1905, fig. 14) and it seems not unlikely that in it 

 are combined the characters of test given by Ritter (for Californian specimens) with the plan of muscle 

 arrangement given by Apstein (1906, PL XXVI, fig. 6) for an antarctic specimen. Such an association 

 of characters has not appeared in any of the specimens examined in this study. 



The External Character of the Test 

 The presence of serration or denticulations on the external surface of the test of some species of 

 salps has been noted by many workers and has in some cases been used as a character for the separa- 

 tion of new species, subspecies or variants; for example, T. democratica var. orientalis Tokioka 

 (1937). S- fnaxima var. tiiberculata (Metcalf, 1918) and S. fusiformis aspera (Ihle, 191 1). Sewell 

 (1926, 1953) discusses the occurrence of such denticulations at some length both in the Salpida and 

 Pyrosomida and concludes that, owing to the complete intergradation between smooth and serrated 

 forms, they are of little importance as characters for systematic differentiation, and attributes 

 the variability of this feature to the great degree of plasticity inherent in the Salpida (see also 

 Metcalf, 1918, p. 5). The evidence of Yount (1958) would seem to confirm this view and in the 

 case of S. fusiformis serrated and smooth forms are listed as the one species. The results of the present 

 study are contrary to the conclusions of Sewell and Yount, at least in so far as they relate to S. fusiformis 

 and its related species. It was noted at the outset without the use of any special technique that two 

 basic test forms could be recognized, one smooth and the other serrated and while this character alone 

 does not allow all the four species to be distinguished it can be combined with other features and used 

 for their identification. 



Detailed examination of the external surface of the test shows, as Stiasny (1926) has described, the 

 presence of areas of thickening which give it a ridged appearance. In species with serrated tests the 

 denticulations occur along these ridges, the arrangement of which in the solitary form in particular 

 follows a definite pattern, the finer detail of which is easily visible when the test is stained with 

 toluidine blue or methylene blue (see p. 5). As the serrations are a feature of the external surface of 

 the test they are subject to wear, and with increased age tend to become smooth, particularly in the 

 aggregate form. Staining, however, allows the remnants of the serrated areas to be detected in 

 specimens which to the casual observer might appear smooth. 



The aggregate form of S. fusiformis (p. 9) has a smooth test devoid of any serrations and so can be 

 distinguished from the three serrated species. It is not possible, however, in the aggregate form to 

 differentiate the serrated species from each other on the basis of the external test, since the pattern of 

 arrangement is less definite and possibly more subject to variation. Of the species, the two which 



