1909] Torrey-Kleeberger: Three Species of Cerianthus. 123 



Accepting the name, the resl of van Beneden's statemenl will 

 hardly hold in view of the facts supplied by C. benedeni. The 

 filaments bearing cnidorages are identical in structure with the 

 filaments riot so furnished, such as those on the mesenteries of He- 

 third and fourth orders. Further, while cnidorages appear most 

 commonly and abundantly on the filaments bunched on the 

 mesenteries of the first and second orders, they are also found 

 commonly, in a less luxuriant development, on many of the 

 scattered filaments of the same mesenteries. They are found only 

 on filaments. Also, these filaments correspond perfectly to the 

 filaments figured by von Heider for C. membranaceus. Since 

 van Beneden insists (p. 33) that acontia differ totally — in posi- 

 tion, morphological value, structure and function — from these 

 filaments in C. membranaceus, the probability that acontia and 

 botrucnides are morphologically comparable vanishes. The single 

 enidorage, not the aggregate of cnidorages into a botrucnide, 

 would appear to be the essential unit of difference between all 

 other species and those which possess it."'' 



C. johnsoni, n. sp. 



The two individuals from which the following Tacts were ob- 

 tained, came from San Pedro Harbor, where they occupied 

 approximately the same habitat as C. bt ned( ni in San Diego Bay. 

 They are much contracted, but were in life obviously much 

 larger than C. aestuari, and considerably larger than our speci- 

 mens of C. benedeni. One specimen is 12 cm. long and almost 

 1 cm. in diameter. The column wall is very thick, leathery and 

 muscular. In the two specimens, the tentacles number : marginals. 

 108, 105; labials. 100, 100. Color notes are wanting. 



The internal structure is represented in fig. 4. The siphono- 

 glyph is narrow, with a seamlike groove opposite, as in C. //< m 

 th ni. The mesenteries agree in number approximately with the 

 labial tentacles. They are arranged in four orders, as in C. In tu - 

 il< ni, though but one pair reaches the vicinity of the terminal 

 pore; and the directives do not extend beyond the lower limit 

 of the siphonoglyph. 



ioSee van Beneden's distribution of larval Cerianthidae between two 

 jroups: Aeontif'erae and Botrnenidae. 



