632 INVEETEBRATA CHAI-. 



of Cynthia, partita Nature lias provided us with an ocular demon- 

 stration of the existence of organ-forming substances. 



OTHER GROUPS OF ASCIDIANS 



We must now take a brief glance at the modifications which the 

 life-history which we have described undergoes in other groups of 

 Ascidians. As already indicated, the other simple Ascidians, so far 

 as they have been examined, have a development which is, cell 

 for cell, identical with that of Cynthia. To this statement, 

 however, certain species (but not all) of the genus Molgula form an 

 exception. 



In Molgula, ampulloides, as described by Damas (1902), the early 

 development is identical with that of Cynthia, but the tail rudiment 

 remains short, and the hinder part of the nerve tube, the notochord, 

 and the tail muscles, undergo degeneration before the larva escapes 

 from its membrane. The notochord and muscles are indeed converted 

 into a mass of fatty cells 'which forms a rounded protuberance at one 

 side of the larva. The larva cannot swim, and tumbles at once to 

 the bottom, where it attaches itself by long finger-like processes of 

 the test. The atrial cavity is formed by a single mid-dorsal involution 

 of the ectoderm, which forks below into right and left pouches ; each 

 of these pouches comes into contact with two evaginations of the 

 pharyngeal wall which are the endodermal portions of the first two 

 gill-slits on either side. 



In the Ascidiae compositae, the larva differs from that of simple 

 Ascidians in the precocious appearance of adult features ; thus the 

 atrial cavity generally attains the form of a single dorsal cavity with 

 median opening, whilst the larva is still free-swimming. The chin 

 region of the larva is already disproportionately large, and the mouth 

 and atrial openings are consequently close together. In some cases, 

 as for instance in Clavelina, the eggs are small ; but in other cases, 

 such as Distaplia, the eggs are larger and more yolky, and the 

 endoderm is in the form of a solid mass of cells like those at the 

 lower pole of the frog's egg. In such cases the archenteric cavity is 

 formed by a separation of these cells from one another, not by an 

 invagination, and there is never an open blastopore. The neuropore 

 never closes, but persists until it becomes converted into the opening 

 of the sub-neural gland. 



In Ascidiae luciae, represented by the single genus Pyrosoma, 

 the eggs are still larger and each individual produces only one egg. 

 The yolk is so massive that it remains largely unsegmented, and its 

 development is, like that of Cephalopoda, of the meroblastic type, in 

 which a germ disc or blastoderm rests on an unsegmented mass of 

 yolk. The test cells, or follicle cells, wander inwards amongst the 

 blastomeres during the progress of segmentation (Fig. 459). Many 

 of them die and degenerate and are used as food, but according to 

 Julin (1912), who has given the latest and most satisfactory account 



