MUSEUM OF COMPARATIVE ZOOLOGY. 295 



ductive organs. On each side of the body the atrium is, as it were, 

 tucked into the space between the branchial and the outer tunic, thus 

 forming two long and broad but very shallow " lateral atria," the 

 cavities of which are continuous with that of the mid-atrium ; these 

 lateral atria are extended upon each side of the branchial sac until 

 they almost meet, and thus cover the whole surface of the sac except 

 a narrow line over and parallel with the endostyle. In a transverse 

 section we should have the test and outer tunic as two concentric 

 circles, and, within these, sections of the branchial and mid-atrial 

 chambers side by side and separated from the outer tunic and from 

 each other by the sinus cavity ; the cross-sections of the lateral atria 

 would be shown as two long parallel-walled diverticula from the mid- 

 atrium, curving around the branchial sac, and almost meeting upon 

 its opposite. side. The outer wall of each lateral atrium is separated 

 from the inner surface of the outer tunic, and the inner wall from the 

 outer surface of the branchial sac by blood sinuses ; but the separa- 

 tion from the branchial sac is not complete, for at certain points this 

 is united to the atrial tunic so that the branchial sinus is divided up 

 into a network of longitudinal and transverse vessels or blood-chan- 

 nels, crossing each other at right angles. In each of the parallelo- 

 grams thus formed the two tunics are absorbed, thus forming a " bran- 

 chial slit," through which the respired water passes from the branchial 

 chamber into the cavity of the lateral atrium, through which it is 

 driven into the mid-atrium, from whence it escapes through the atrial 

 aperture. It is difficult to give a clear account of the relations of the 

 branchial sac and atrium without the aid of diagrams, but an illustra- 

 tion may help to a conception of this somewhat complicated subject. 

 In the middle of a long glass tube blow a bulb to represent the stom- 

 ach, and enlarge one end of the tube to represent the branchial sac, and 

 the other to represent the mid-atrium ; the small tube uniting the 

 branchial sac to the stomach will represent the oesophagus, and that 

 which connects the stomach to the atrium will represent the intestine. 

 Now bend the intestine around the stomach so that the branchial 

 sac and atrium shall lie side by side, flatten out the latter, and 

 wrap it around the branchial sac, and the whole will form a pretty 

 correct model of these organs in a typical Tunicate. This illustration 

 is open to one objection, inasmuch as it seems to imply that the bran- 

 chial sac and atrium are serially homologous, while, in fact, the latter is 



