Age -Determination, Growth and Symmetry 31 



fibres in the body-wall. In Chelyosoma the rigidity of the yellow outer 

 layer of the test tends to prevent such general movement, great flexibility 

 being present only along the lines where the yellow layer is lacking. It 

 is natural, therefore, to have the muscles more or less restricted in extent, 

 and passing on the whole from plate to plate perpendicularly to the 

 margins. Expansion and contraction can take place only between the 

 plates, the disk as a whole being very convex in the expanded condition 

 and flat in the contracted. In contraction the plates are separated by 

 deep, narrow grooves. The attached part of the test has no yellow layer 

 deposited on its surface, and its enlargement presents no dif^culty. 

 The mode of enlargement of the remaining part of the test, namely that 

 between the disk and the attached surface, is similar to that of the disk. 

 This part shows a variable network of lines, but no definite grooves like 

 those between the plates of the disk, owing to the absence of intrinsic 

 muscles connecting the areas. Along these lines the yellow layer is 

 absent, and enlargement, therefore, is possible. Near the disk these 

 lines form irregular plates, which are indistinctly in series with those 

 of the disk. As the grooves separating these plates are very superficial, 

 the formation of lines of growth would be unlikely to occur, and in fact 

 none were found. 



The formation of the plates is based upon the apertures. As is so 

 frequently the case in related genera of Ascidians, the apertures of 

 Chelyosoma are surrounded each by six lobes of the test and are borne 

 at the extremity of more or less distinct siphons. It is natural, therefore, 

 for the yellow layer around each orifice to be divided into six plates 

 corresponding to the lobes and extending to the base of the siphon. The 

 number of plates around each orifice is not absolutely constant, Bancroft 

 having observed both five and seven plates around the atrial orifice in 

 C. productum. In C. macleayanum we have failed to observe any varia- 

 tion from the typical number. Beyond the siphonal field there is a single 

 series of plates for each aperture, which we may call the extrasiphonal, 

 those for the two apertures fusing together more or less perfectly into 

 one series, the margin of which is the edge of the disk, elliptical in outline 

 with the aperture form.ing the foci. In the present species this series 

 shows five plates for the oral and four for the atrial aperture, one plate 

 (the central) being common to the two apertures (See PI. I, fig. 2). The 

 number and arrangement of the plates in this series are more variable 

 than with those of the siphonal series. Bancroft has divided them into 

 central (between the siphons), and peripheral (next the margin of the 

 disk) . The latter are divided into lateral (touching the central plate) and 

 anterior and posterior (touching only the siphonal plates). We have 

 observed an additional lateral plate on one or both sides, and only one 



