440 
above (fig. 5) its anterior margin shows two transverse impressious, 
indicating that the joint in reality is formed by the fusion of three 
joints. The eighth, ninth and tenth joints are all subequal in 
length and together only a little longer than the seventh; the 
penultimate joint is a little shorter than the two preceding ones 
together; the terminal joint is only a little shorter than the five 
proximal joints together, with the end blunt and irregularly cut olf. 
On the distal angle of the penultimate joint is inserted an olfactory 
seta, the length of which is two-fifths of the whole antennula; this 
seta is very thick, but in the specimen examined a part of it beyond 
the middle seems to be collapsed in both antennulæ, and the por¬ 
tion in question is not similar in the two setæ (comp. fig. 4 with 
fig. 5). The setæ on the three proximal joints are very short and 
apparently few in number (some may be overlooked or broken off); 
the fifth joint possesses five long, robust setæ which taper gradually 
to the acute end, and each of nearly all the other joints has a 
single or two setæ of the same shape; the last joint has five very 
thin terminal hairs. On the seventh, eighth, ninth, tenth and 
twelfth joints are found several setæ of another shape: these setæ 
are considerably thicker in proportion to their length, taper gradually 
towards the end which is somewhat thickened and generally (perhaps 
always) emarginate or slightly cleft. The seventh joint has two 
short and two rather long setæ, the ninth joint one rather long and 
proximally very thick seta and one shorter and less thickened seta, 
each of the three other joints only one rather short seta of this 
kind. In my opinion, these setæ are scarcely supplementary olfactory 
organs, but it must be added that the state of preservation and 
the scantiness of material checked the study. — I do not venture 
to attempt an interpretation of all the joints of the antennulæ ac- 
cording to the formula of Giesbrecht, showing how the number has 
originated by fusion of certain groups of twenty-five joints, but 1 
think that the joints in the proximal half of the antennulæ must be 
interpreted in the following way: 
