Guff// Marine Laboratory, St. Andrews. 25 



fig. 38 only the section of a single tube on each side is indi- 

 cated. This interpretation shows certain differences from 

 the arrangement in Bispira. Brunotte's view that the walls 

 of these tubes (individual folds) are specially arranged holds 

 only good in Bispira, so far as it refers to folds of the 

 appendicular duct posteriorly (PI. II. fig. 11, tg.). The 

 author is inclined to think that these thoracic segmental 

 organs represent the series found in other forms, and are 

 probably homologous with the longitudinal canal in Lanice. 

 Tlie thoracic glands (anterior nephridia) in Bispira and 

 other Sabellids follow a different arrangement from those in 

 the Serpulids, e.g. Pumatocerus, which have tlteir widest part 

 anteriorly and diminish in their progress backward to a 

 blind end. In longitudinal section these glands fill the 

 coelomic spaces of the first two segments in Bispira, which 

 thus agrees with Branchi omnia as described by Brunotte, 

 though their convolutions would appear to be larger, such 

 depending to a certain extent on the degree of contraction 

 or expansion. In the serial (transverse) sections from the 

 front the first trace observed is a small tube with pigmented 

 walls situated about the level of the upper arch of the gullet, 

 between the approximated dorsal and ventral longitudinal 

 muscles, and it is imbedded in muscular fibres stretching 

 from the gullet to the body-wall. Such represents the 

 anterior duct of each side, thus corresponding to the arrange- 

 ment in the Serpulids. The thoracic gland increases 

 gradually in size and passes downward to the exterior of the 

 oesophagus, resting on a plate of muscle passing outward to 

 the wall and cutting off a coelomic space above it on each 

 side. Here the small tube has fixed to it a loop of vesicular 

 and cellulo-granular tissue which seems akin to the chlora- 

 gogeuous investment of the gut, the cells and vesicles 

 hanging on a thin mesenterial tissue in groups (PI. II. 

 fig. 11, chl.). The structure of the gland in section is similar 

 to that in the Serpulids, but the walls are, perhaps, less 

 massive than in Pomatocerus, though of considerable thick- 

 ness, the tough external layer having muscular fibres within 

 it and the epithelial layer being largely developed. With 

 the increase of the coelomic space the gland on each side 

 moves downward and the cellular loop (really a tube) 

 enlarges, and the sections of the gland lie within the ring 

 of this tissue. Then sections of two glandular tubes 

 appear, as if the organ had become bifid, both connected 

 with the granular cellular tissue, the vesicles and cells 

 projecting into the ring from the limiting membrane ex- 

 ternally, and they form a thicker and more definite layer. 



