288 bulletin: museum of comparative zoology. 



circular than in the cotypes of that species from Brazil. But none of 

 these differences is anything more than of degree, and it is unlikely 

 that they have any real taxonomic significance. In a typical A. 

 planispina, each radial shield is 1.4 mm. long by 1 mm. wide, while in 

 the Tortugas specimen, each shield is 1.8 mm. long by 1 mm. wide. 

 In typical A. planispina there are only 10-15 disk-scales to a square 

 mm. while in the Tortugas specimen there are 20-25. In life, the 

 Tortugas specimen had the disk gray with a yellowish tinge and the 

 arms yellowish white; a few irregularly scattered upper arm -plates 

 are partially dusky purplish, but there is not nearly so much dark 

 coloring as in typical A. planispina. 



Amphiodia rhabdota, sp. nov. 

 pa/35co7-6s = striped, in reference to the longitudinal line on the arms. 



Plate 8, fig. 4. 



Holotyjje.— M. C. Z. 4,216. Florida: Dry Tortugas, Bush Key 

 reef-flat, June 18, 1917, in mud and eel-grass roots, in 2-3 ft. of water. 

 Carnegie Expedition. H. L. Clark coll. 



Disk 6 mm. in diameter; arms 40^5 mm. long. Very similar to 

 A. planisjnna but the disk-scaling is much finer, the radial shields are 

 long and the upper arm-plates are different. Orally, the two species 

 are identical save in coloration. In A. rhabdota, there are about 50 

 disk-scales to each square mm. The radial shields are nearly three 

 times as long as wide (1.7 X .6 mm.) and their proximal ends are a 

 little separated. The upper arm-plates are broadly oval, rounded 

 and not squarish on the lateral margins, and are only twice as wide as 

 long. In coloration, the disk is gray and the arms yellowish white 

 as in yl. planispina, but the irregular dusky markings on the arm are 

 yellowish and very faint, while there is a distinct, narrow, dusky 

 longitudinal stripe on the upper arm-surface; this line is occasionally 

 broken or interrupted. Nearly all the under arm-plates have a median 

 longitudinal dash of brownish red and these dashes tend to form an 

 intermittent narrow stripe. 



It does not seem possible to consider this specimen an immature 

 example of A. planispina, for the changes of growth do not decrease 

 disk-scales, and no case is known of radial shields becoming so much 

 shorter and wider with age. Of course there is a possibility that a 

 large series of Amphiodias from the West Indian region, will show A. 



