8 EISEN 



surface of the body. There are no muscular bands connecting the 

 base of the cushions with its periphery. The sperm-duct never pene- 

 trates the bulbs or cushions but opens close to and independently of 

 them. Exterior to the cushions there are numerous muscles connecting 

 the body wall immediately surrounding the pore with other parts of 

 the same somite. 



The Lumbricillid bulb is always single and covered with a strong 

 muscular layer, which however never penetrates down between the 

 cells of the bulb. There are generally two or three distinct sets of 

 glandular cells in the bulb. Some of these open in the lower part of 

 the sperm-duct, or rather in a narrow groove in the elongation of the 

 sperm-duct. Others open on the free surface of the bulb, either irregu- 

 larly or in narrow circular fields, bunched into fascicles. The sperm- 

 duct penetrates one side of the bulb. In Bryodrilus the gland which 

 opens in the extension of the sperm-duct is covered with a thin cushion 

 of muscular strands, forming a bulb within a bulb. 



Structure of the atrium and its glands. The structure of the 

 enlargement of the sperm-duct which I have designated as atrium is a 

 complicated one, especially in Mcsenchytraus. In the subfamilies of 

 Lumbricillinae and Enchytraeinaj the sperm-duct continues to the pore, 

 even through its passage through penial bulb, without any enlargement, 

 and without being joined by any atrial or accessory glands. Any ref- 

 erence to the finer structure of the sperm-duct proper in these two sub- 

 families is therefore not necessary. But in Mesenchytrceus the struc- 

 ture is often so complicated and so varied that it generally furnishes im- 

 portant characteristics of the species. In many species there exists an 

 atrial enlargement just outside of the penial bulb, while many species 

 possess also another enlargement inside the penial bulb, close to the 

 penial pore. For the former I have retained the name ' atrium,' for the 

 latter l penial chamber/ Both these enlargements may be connected 

 with various kinds of glandular cells. These cells are either single or, 

 more frequently, grouped in fascicles in the same manner as the septal 

 glands. All the various glands in the family resemble one another in 

 that the respective cells open independently of each other through a 

 long and narrow duct. In no instance is there a common lumen for 

 the various cells, though they may be grouped together in fascicles, in 

 which the long and exceedingly narrow ducts run parallel to each other 

 for some considerable distance. This is especially the case with the 

 atrial glands. These glands occur generally in fascicles, which lie 

 free in the coelomic cavity, but send their fine, thread-like ducts into 

 the atrium of the sperm-duct. In many species the ducts of the fasci- 



