82 Dr. 8. Lovén on a remarkable Sponge 
dermal layer. Its inferior end is thickened into a dilated base, 
from which the fine root-fibres spread in branches, forming 
loops, and having attached to them numerous grains of sand, 
spicules of sponges, and Foraminifera. 
A closer examination has given the following results. The 
dermal layer of the stem is thin, but tough, and may be drawn 
off in long pieces. It then shows a transparent uncoloured 
protoplasma, full of small yellowish granular corpuscles, with 
or without larger granules (fig. 4). In this parenchyma is 
imbedded (fig. 5) a felt of very small siliceous spicules, spindle- 
shaped, not inflated in the middle, furnished with a central 
canal (fig. 6). When measured, these were found to be from 
0-1 millim. in length and 0:0018 millim. in thickness to 
0°08 millim. in length and 0-002 millim. in thickness ; the 
mean length was 0:08 millim., and the relation between length 
and thickness in one as 100: 3°6, in another as 100: 1°8, 
the mean of eight measurements as 100: 2°76. The granules 
of the parenchyma are more discernible if prepared with gly- 
cerine, while the spicules are more distinct in pesads balsam. 
Within the dermal layer the stem is made up of closely 
packed spicules, held together by a relatively small quantity 
of parenchyma (fig. 7). At first sight it seems as if the stem 
were composed of very long, rather spiral filaments; but a 
closer examination shows the spicules to be very short, but 
disposed in strings ; so that the whole has the aspect presented 
in fig. 8. The spicules are all of the same type: they are 
spindle-shaped needles (figs. 9, 10, 11), having near the middle 
a slight but distinct globular inflation or nodule, and tapering 
towards either end from that point, not in a straight line, but 
forming together a very obtuse angle. It is owing to this 
peculiarity that the needles, united in rows, produce the 
slightly spiral structure of the stem. Every needle ends in a 
fine but rounded. point (fig. 12). They are more or less round. 
The layers of which they consist are not to be discerned ; only 
the exterior one appears in the transverse section (fig. 13) as a 
very thin ring, They have a fine central canal, which, if the 
needle is not broken, is closed at the point. When the infla- 
tion in the middle is not larger than is shown in figs. 9,10,11, 
the central canal goes through it without branching; but if 
the nodule has increased a little more in two opposite direc- 
tions (as is shown in fig. 14), which is very seldom the case, 
two fine but distinct transverse canals are seen to go off cross- 
wise from the central canal into its nodule or inflation. I 
have not observed this formation of secondary canals in the 
middle nodule carried further than shown in fig. 14; it is an 
incipient branching, and appears also in other parts of the 
