LIVING BRACHIOPODA. 329 



there are certain sedentary forms as well as a few errantian forms in which unjointed and 

 barbed setae are found ; in Sal)ellifonnis, for example, the setae are unjointed, hair-like, 

 and fringed with short setellae. Sphaerodorum is exceptional among the errantian anne- 

 lids in having unjointed setae ; in the Cajntellidde, also, the setae are unjointed. 



In the errantian Brachioj^oda, Glottidia and Lingula, as we have already seen, the 

 setae act as locomotive organs, swinging with the same rhythmic motions that characterize 

 the cliaetopods. The setae were primarily locomotor in function and persist through 

 inheritance, a secondary function being acquired in tlie LinguJidae in forming the lateral 

 barriers to protect the pallial cavity from the ingress of sand and mud. In other groups, 

 a secondary acquisition has pro)>ably been established in furnishing a nidus for diatoms 

 and other foreign growths which they sustain, often in rich abundance, for the benefit of 

 the animal in furnishing, or attracting, a food supply. In all brachiopods, they undoubt- 

 edly act as a sensory apparatus for tactile impressions. 



The setae in all brachiopods thus far observed are marked by transverse lines, or 

 joints, which are shouldered and whicli indicate successive increments of growth. Even 

 in the almost aborted setae of Dallina the transverse lines are easily detected. Fine par- 

 allel lines also mark these hairs, and their appearance has been happily compared to the 

 stem of an Eqviisetum. 



In examining the setae of different genera more in detail, we shall find that in the 

 unattached group, LhujuUdae, they are most highly developed. Furthermore, the setae 

 are grouped in definite clusters. On the anterior median line of the animal is a group of 

 setae in which the middle ones are the longest, the setae on each side diminishing rapidly 

 in length. This may be called the median cluster. The setae on each side of this cluster 

 increase gradually or aljruptly in length according to the species, becoming longest at the 

 round angle of tlie shell where the anterior and lateral borders of the shell meet. These 

 stand out at an angle from the medial] line of the body, right and left, and may be called 

 the anterior clusters. Fi-om these clusters on the side, the setae very slowly diminish in 

 length till they reacli that part of the shell where it begins to taper rapidly toward the 

 peduncular end; at this point they immediately become much longer, forming a very 

 sharply defined cluster. These groups may be called the right and left posterior clusters. 

 The setae between the anterior and posterior clusters may be designated as the lateral 

 setae (40: 1). These clusters, with the anterior clusters, are peculiarly active when the 

 creature is crawling over the sand or burrowing into it head foremost. Behind the pos- 

 terior clusters, the setae are less numerous, and toward the j^eduncular end become more 

 filiform, and are often bent and broken. They appear to be functionless and are becoming 

 rudimentary, presenting a condition not unlike that seen in certain Testicardine forms. 

 The proportionate length of the setae, and the outline made by the median and anterior 



