The Retina and Optic Ganglia in Decapods, especially in Astacus. 19 
half-plate and not due to the penetration of fibres from one half- 
plate into another. 
At the base of each half-plate where the fibres meet the reti- 
nular celi (Fig. 23), the pigmented substance of the celi extends as 
a blunt process into the piate. In this way, transverse pigment 
bands are produced on the four surfaces of the rhabdome. Since 
each band marks the attached face of a half-plate and since antero- 
posterior plates alternate with dorsoventral ones, the well-known 
alternation of the pigment bands on adj acent sides of the rhabdome 
and their coincidence on opposite sides are easily iinderstood. 
The precise nature of the fibres is by no means easily deter- 
mined. That they are not purely artificial products of Golgi's me- 
thod is shown, I believe, in two ways: first, they occur only in the 
rhabdomes and bave a definite arrangement conformable with the 
structure of these bodies, conditions that would not be expected if 
they were simply artificial products, and, secondly, they can be 
demonstrated in some crustaceans, as, for instance, in Porcellio and 
Serolis without the use of this method^. In deciding on their 
nature, one structural feature that they present is of considerable 
importance; in ordinary sections it is not uncommon to see 
streaks free from pigment extending from the base of the fibres in 
the rhabdome through the pigmented substance of the retinular celi 
to its fibrillar axis as though the fibres and the axis were in direct 
anatomical continuity, and in preparations made by Golgi's method, 
the Silver salts often take this course, coloring the connecting streaks 
and the fibrillar axis as well as the fibres themselves. This favors 
the view that the fibres represent the distai terminations of the 
constituents of the fibrillar axis, and throws some light on the follow- 
ing peculiarity of the rhabdome, as contrasted with a nerve fibre, 
in their relations to beat. When a perfectly fresh nerve fibre is 
heated to 50° C, it contracts about Y4 or 1/5 i^s originai length; 
when a fresh rhabdome is similarly treated, it becomes thinuer and 
^ The fibrous structure of the rhabdome has, in fact, already been recorded 
in a number of arthropods. It is probable that what Max Schultze (68, pag. 14, 
PI. 1 Fig. 12) described and figured as lamellae in the plates of the rhabdome 
in Astacus were really fine fibres seen from the side. Patten (86, pag. 629), 
however, was the first that recognised fibres as such in the crustacean rhab- 
dome. Similar structures have been observed by Watase (90, pag. 291) and 
myself (91, pag. 117) in Serolis, by Watase (90, pag. 303) in Limulus , and by 
Willem (92, pag. LXXX etc.) in Lithohius and Folyxenus. 
2* 
