u6 MCMURRICH. [VOL. i. 



mesenteries are usually present, nevertheless, in a number of 

 cases, a reduction to a single pair has been observed, not only 

 in the Halcampidas, but also in other less primitive families, 

 and more especially in the Sagartidae. 



This condition, which is associated with a corresponding 

 reduction of the number of siphonoglyphes, was observed by 

 Thorell ('58) and has been studied more recently by A. F. 

 Dixon ('88), Carlgren ('93), and Parker ('97). I have made some 

 observations on monoglyphic specimens of Metridium margi- 

 natum, but Parker's observations on the same species have been 

 so much more extensive that a record of my results seems 

 unnecessary, as they throw no additional light on the signifi- 

 cance of the abnormality. 



The reverse condition, an increase in the number of the 

 directives beyond the normal, is apparently much less frequent, 

 and has, up to the present, been described only by G. Y. and 

 A. F. Dixon ('89) in Bunodes thallia, and by Parker ('97) in 

 Metridium marginatum. Parker found three siphonoglyphes 

 and three pairs of directives in only one out of one hundred 

 and thirty-one specimens ; but in four specimens of Bunodes 

 thallia the Dixons found that one had two pairs of directives, 

 one three pairs, one four pairs, and one only one pair. In the 

 Bunodes as in the Metridium the siphonoglyphes corresponded 

 in number with the directives, and furthermore it is to be 

 noted that an irregular arrangement of the mesenteries was 

 also found in all the specimens which presented what may be 

 termed, following Parker's nomenclature, a polyglyphic condi- 

 tion. An increase in the number of siphonoglyphes has also 

 been found in two other species and probably denotes a cor- 

 responding increase of directives ; the Dixons ('91) describe 

 the occurrence of three grooves in a specimen of Metridium 

 dianthus, and Haddon and Shackleton ('93) state that from two 

 to seven (!) siphonoglyphes occur in different specimens of 

 Conclylactis Ramsayi. 



I have recently had the opportunity of examining a species 

 in which the number of the directives seems to be as a rule 

 increased above the normal. This is the form originally 

 described by Verrill ('83) as Sagartia spongicola, several speci- 



