56 
Canada balsam it is seen that there is really not a trace of scales 
in the middle of the disk, while only a very few, small scales are 
found covering the inner end of the radial shields. 
Another character, the length of the arms, which is, according 
to Sars, 6—7 times the diameter of disk in borealis , while in 
securigera it is as much as 12—15 times the diameter of disk, 
Fig. 3. a—c. Amphiura borealis ; d—f. A. securigera. 
a. and d. ventral side, b. and e. dorsal side, c. and f. part of arm from the dorsal 
side, showing the widened armspines. u /i. 
gives also a good distinguishing character — when well preserved 
specimens are dealt with; unfortunately, these species appear to 
be very fragile, the arms very easily breaking, so that this char¬ 
acter is practically of very little use. One small tentacle scale may 
occur fairly regularly farther out on the arms of securigera ; in 
borealis I have never seen a trace thereof. In the oral parts and 
ventral plates (Figs. 3, a, d) no distinet difference exists between 
the two species, whereas the dorsal plates afford a good disting¬ 
uishing character, being more rounded in borealis than in securi¬ 
gera, and also contiguous in borealis , while they are distinetly 
separated from one another in securigera. Finally the development 
of the axeshape of the second armspine is much stronger in secu¬ 
rigera than in borealis (Figs. 3, c, f). 
