55 
1917-18.] Respiratory Mechanism of the Shore-Crab. 
later Garstang (3) opposed this contention, while he himself described a 
similar reversal in Corystes, Atelecyclus, and Portumnus (5). He, however, 
suggested that the phenomenon did not exist in such decapods as Cancer 
and Carcinus, and states positively ( 5 ), in the case of the latter, “ I have 
found no indications of a reversal of the respiratory currents in the latter 
species” (Carcinus). His observation in this respect has been shown to 
be erroneous by Bohn (1), who seems to have proved that reversal is 
common to all decapods. My results confirm those of Bohn. They show 
that reversal occurs under all conditions, and that it is more frequent when 
the crab is buried or when a stronger ink suspension is used. On such 
meagre evidence I do not attempt to criticise this last authority, who argues 
that reversal is due to fatigue of the muscles which cause the forward 
direction of the scaphognathite. 
I have also endeavoured to determine the direction of the respiratory cur- 
rents in the branchial chambers. While investigating the cause of the “ sand 
eddies ” previously mentioned, I came to regard the position and relations of 
the gills as of first importance in the determination of this direction. 
Claus (2) describes a backward flow due to the action of the maxillipede 
flabella. Pearson (8) agrees with Claus as regards direction, but fixes the 
cause to the presence of the “ branchial ridge.” Bohn, on the other hand, 
maintains that the direction is postero-anterior. The differences in opinion 
arise mainly in reference to the extent of the inhalent apertures ; and since 
the two former workers believe that the whole sub-branchial cleft does not 
admit of a current, the aperture round the chela is considered to be the 
main inlet. This being situated far anteriorly, the backward flow becomes 
a necessity in order to thoroughly bathe the posterior gills. From my 
observations on the parts in question, it seems quite plain that the sub- 
branchial inlets will, to a large extent, predetermine the flow of the inhaled 
currents. The direction of these inlets veer from postero-anterior to almost 
latero-medial (from behind forwards), tending, therefore, towards the pre- 
branchial chamber dr outlet — see figs. 1 and 6. Thus it is not difficult to 
allow that the inhaled currents must also follow this direction ; additional 
evidence of this supposition being correct is supplied by Experiment III. 
In connection with the above, there is another interesting feature in 
the radial arrangement of the gills and the direction of each of the sub- 
branchial inlets. These are so related that the gills lie more or less con- 
stantly at an acute angle to the direction of the inhaled currents, which, 
meeting with the opposition of the branchial septa and the gill lamellae, 
can only pass upwards between the individual lamellae to the roomy roof 
of the branchial chamber. Some will no doubt flow along the “ tunnel ” 
