254 CEYLON PEARL OYSTER REPORT. 



macerated or much contracted gills the large oblong cells become loosened from their 

 attachment to the filament and become bent up in a crescentic form with their 

 convexities outwards. In this condition, when the cilia remain attached to them, 

 they mav very easily be mistaken for ciliated discs, and it was only after studying 

 well-preserved preparations with the highest powers of the microscope that I discovered 

 the real state of the case. As far as I am aware, very large elongated cells of this 

 shape bearing the lateral cilia have not been described before, and they seem to be 

 peculiar to Jousseaumia. It is, as I have said, possible to regard the longer cilia on 

 the frontal edges as latero-frontal, and in that case the very long stiffer cilia succeeding 

 them would be lateral cilia, and the long fine interlocking cilia borne by the brick- 

 shaped cells might be regarded as occupying the position of and being homologous 

 with ciliated discs. On this view the gill of Jousseaumia would have to be regarded 

 as a primitive form of filibranch gill, in which the interlocking cilia are arranged 

 in continuous lines and are not differentiated into isolated ciliated discs. But this 

 view is hardly tenable. The gills of Jousseaumia are not filibranch, for they have well- 

 developed interfilamentar junctions. Moreover, the interlocking cilia, in addition to 

 their being arranged in longitudinal lines and not in groups, are actually finer and 

 longer than the fronto-lateral cilia, and lack the short, stiff brush-like character of the 

 cilia of true ciliated discs. The fact remains, however, that they interlock, and that 

 there is therefore a ciliary union in addition to an organic union between the filaments 

 of Jousseaumia. It seems to me probable, however, that the physiological role of the 

 interlocking cilia is rather to form a barrier preventing solid particles from passing 

 between the filaments than to give mutual support to the filaments, and this view is 

 supported by their extreme fineness, while the coarser latero-frontal cilia projecting 

 from the corners of the frontal edges are evidently effective in sweeping solid particles 

 over the surfaces of the gills towards the labial palps and mouth. 



The interfilamentar junctions are arranged in regular rows. In most specimens 

 there are three such rows in the anterior part of the direct lamella and one or two 

 rows in the reflected lamella. As has been stated, these junctions are non-vascular 

 and are formed as secondary outgrowths from the filaments, bridging across the 

 fenestra? at regular intervals. As may be seen in figs. 17 and 21, these interfila- 

 mentar junctions are curved bars, continuous with the chitinoid lining of the central 

 cavity of the filament, but the junctions themselves are solid, and as they are only 

 clothed by a very thin protoplasmic sheath, they do not establish any vascular 

 connection between adjacent filaments. As seen in section, the interlamellar edge of 

 each filament appears to be prolonged to form a pair of bars which curve round to 

 unite with corresponding outgrowths from the adjacent filaments. The lower part 

 of fig. 22 shows the interfilamentar junctions as viewed in optical section under a 

 very high power of the microscope. As a rule the interfilamentar bars are single, 

 but occasionally they are double, as shown in the middle of the figure. The chief 

 point of interest is that the bars are clearly shown to be formed by the agency 



