258 Vnivcrsity of California Puhlicatioiis in Zoology [Vol. 18 



This hypothetical distribution of coi'ks and iron weights exactly 

 parallels the distribution of the individuals of the two generations of 

 Salpa democratica as demonstrated by tables 4 to 6. The conclusion, 

 therefore, seems unescapable that each aggregate salpa, after being 

 pushed to the exterior of the solitary salpa, remains attached to its 

 predecessor and that this chain of salpae also remains attached to the 

 solitary salpa as in other species of the genus. 



3. Evidence of the Existence op Protruding Chains 



The conclusion being reached that protruding chains of Salpa 

 democratica are of normal occurrence, it is well before assuming its 

 truth, to consider whatever other evidence there may be. To this end 

 the literature has been searched without finding a single unequivocable 

 statement. Brooks (1893, pi. xliii) has published a drawing of what 

 he calls: "x^art of a fully grown chain of Salpa democratica." It 

 consists of six salpae attached together as described on page 244, but 

 the magnification is not given and, while it is probable from the con- 

 dition of the test that Brooks had a portion of a protruded chain 

 before him, it is possible that by "fully grown chain" he means the 

 final positions assumed by the salpae in the terminal portion of the 

 stolon. Again Herdman (1889, p. 58), after refering to a cut of 

 the posterior end of the solitary form of S. democratica, says : ' ' After 

 the solitary Salpa has become fully developed, the chain produced 

 by the stolon is set free in sections, each section being composed of a 

 number of aggregated Salpae at about the same stage of develop- 

 ment. ' ' This statement sounds definite and explicit, but it is not clear 

 that it refers to S. democratica in the first place nor to the freeing 

 of the salpae from the mantle cavity of the solitary form in the second 

 place: by "set free in sections" Herdman may refer to the periodic 

 manner in which the stolon segments (see page 244). 



But, Herdman 's statement is corroborated by that of Agassiz 

 (1866, p. 20) who, in describing S. cahotti — probably a large variety 

 of S. democratica — says: "The young Salpae are not uniformly 

 developed in proportion to their distance from the base of the tube 

 [stolon]. Sections of the tube are equally advanced and we find 

 generally, three such portions unequally developed. . . . The base of 

 the geminiferous tube is simply slightly corrugated, next comes a 

 section in which we find two rows of slight elevations, and finally the 

 most advanced part of the chain where the rudimentary Salpae are 



