BURROWING BARNACLE — TOMLINSON AND NEWMAN 525 



(counts for one specimen), and increase in the posterior cirri. The 

 rami are armed with long setae along their lesser curvature (fig. 10). 

 The number and arrangement of these setae repeats itself on each 

 article. A single seta occurs on every second to fifth articulation along 

 the greater curvature of each terminal cirrus. 



The posterior end of the body supports a pair of uniramous caudal 

 appendages of two distinct segments each (fig. 9). The proximal seg- 

 ment bears a faint indentation suggesting a third articulation. The 

 total length of the caudal appendage is about half the length of the 

 pedicle of the posterior terminal cirrus. 



Male (fig. 4) : The degenerate male appears as a small simple sac 

 with antennules for attachment. It is unique in that it possesses a 

 stalk, often long and attenuated, between the antennules and the body 

 proper. The stalk arises from a T-shaped connection with the two 

 normal-appearing antennules and terminates in an annulated attach- 

 ment to the body. The presence of a penis could not be confirmed. 

 The male attached to the holotype female measured 0.91 mm. in over- 

 all length, and 0.34 mm. in body length. Usually one male was re- 

 covered in or near a pocket on the right side of the female (fig. 3) 

 although two males attached to a single female were observed. 



Phylogeny: Of the known forms of the Lithoglyptidae, Lithoglyptes 

 spinatus from Jamaica is most closely related to L. ampulla from the 

 Java Sea. Both of these forms are superficially similar and both 

 inhabit corals. L. spinatus can be separated from L. ampulla by its 

 proportionately larger aperture, its more numerous small hooks and 

 spinulcs on the surface of the mantle, and by the lesser number of 

 articles composing the rami of the mouth cirri and the caudal append- 

 ages. L. ampulla is twice as large as L. spinatus. 



Armor in the form of large apertural hooks and spines would appear 

 to be a primitive characteristic in the Acrothoracica. The arrange- 

 ment of these structures occurs in a strikingly similar pattern in certain 

 members of the Lithoglyptidae, Balanodytidae, and the Kochlorinidae. 

 It is highly unlikely that such similar patterns could have arisen inde- 

 pendently from the more naked members of these groups. Further- 

 more the most specialized (reduced) acrothoracicans have the most 

 specialized habitats (e.g., Trypetesa within snail shells). The ability 

 to attach by a disc and cement gland is also probably a primitive trait. 

 The location and nature of this gland as seen in Lithoglyptes 

 spinatus makes it highly probable that it is homologous to the cement 

 gland and method of attachment seen in the thoracican Cirripedia. 

 Acrothoracica without this gland should thus be considered as more 

 highly modified forms. 



It follows then that the Lithoglyptidae is the most primitive family 

 of the Acrothoracica because of the heavy armament, the possession of 

 cement glands, the large number of terminal cirri, and the caudal 



