GIGANTOCYPRIS MOLLERI 215 



of the anterior pericardium. As they emerge into the main pericardium they twist over 

 each other so that the most anterior, the eighth, comes to attach most posteriorly and 

 vice versa. 



The extent of these dorsal longitudinal muscles is shown clearly in Fig. 9 and in 

 Plate XLII. The attachments of the outer muscles to the pericardial floor are shown in 

 Fig. 7 b where the body wall, covering and so forming the roof of the afferent canal, 

 has been removed to expose its floor. 



In Doloria (1931) I described this lateral group of dorsal longitudinal muscles and 

 stated that at their lower ends they attached directly to the ectoderm (p. 452). This 

 statement, I am certain now, is incorrect. The small size of Doloria, but more par- 

 ticularly the fact that I was using ordinary paraffin sections, made it very difficult to 

 settle where the muscles actually attached. I have re-examined the sections and find that 

 the pericardial floor is close up against the ectoderm, probably as a result of technique 

 employed. In the celloidin slices of Gigantocypris there is no distortion of the tissues, 

 and I can trace the lower termination of these muscles without any doubt as a mat of 

 tendo-fibrils in the pericardial floor itself (Plate XLII, figs. 5-8). 



From the hinder limit of the caudal furca a group of three broad muscles, the peri- 

 cardio-furcal muscles, radiate forwards, upwards and outwards (Fig. 9). The lowest 

 of these attaches to the e sclerite system associated with the second trunk limb, just where 

 the pericardial floor attaches to the body wall. The upper attaches to the pericardial 

 floor at the lower end of the afferent canal, its tendo-fibrils mingling with those of the 

 outer dorsal longitudinal muscles (Fig. yb). The middle muscle attaches in an inter- 

 mediate zone, but the tendo-fibrils at the upper ends of all three muscles spread out 

 and intermingle. More important, however, is the fact that they intermingle with the 

 mat of tendo-fibrils forming the lower ends of the outer series of dorsal longitudinal 

 muscles. 



This tendo-fibrillar zone at the bottom of the afferent pericardial canal is thus held 

 in position from three directions. Anteriorly it is attached directly to the e sclerite 

 and so to the main skeletal system. Postero-medially it is connected to the caudal furca. 

 Dorsally it is connected by the dorsal longitudinal muscles direct to the shell at the sides 

 of the main pericardium. This dorsal connexion, however, is in effect dorso-medial, 

 for the upper attachments are much nearer the middle line than are the lower, as can be 

 seen from the series of sections shown in Plate XLII. 



Contractions now, of the pericardio-furcal muscles, will pull inwards, downwards 

 and backwards the floor of the lower end of the afferent pericardial canal. The back 

 pull will be resisted by the skeletal system through the e sclerite. The inward pull will 

 open up the entrance to the canal from the main body cavity. The downward pull will be 

 counteracted by the upward pull of the lateral series of dorsal longitudinal muscles. 

 This series, however, lying on the floor of the afferent channel and curving inwards as 

 it passes upwards will, by its contraction, at the same time tend to depress inwards the 

 floor of the canal over the whole of its length. These two sets of muscles, therefore, the 

 pericardio-furcal muscles together with the lateral series of dorsal longitudinal muscles, 



