Die B. GRASSI AND A. SANDIAS. 
much the smaller when the number of antennal joints has 
reached twelve, with the fifth joint already pilose, or with the 
third and fourth alone bare. I may note here that the appear- 
ance of intestinal Protozoa (vide infra) coincides -with this 
last stadium, though Joenia is confined to the large-headed 
larvee, whereas Monocercomonas is common both to these 
and the small-headed forms. 
The development of the four secondary Malpighian tubules 
proceeds in such a way that each is placed midway between a 
pair of the primary tubules; that is to say, the latter are 
equidistant from each other, and the secondary tubules are 
intercalated at equal distances between them, so as to produce a 
series of alternate large (primary) and small (secondary) 
tubules. 
Their mode of development is shortly as follows: they 
spring from the proctodeum at its junction with the mid-gut, 
exactly at the same level of the primary series. I have been 
unable to detect any special layer of embryonic tissue destined 
to give origin to them, and they may therefore be regarded as 
a direct derivation from the proctodeal epithelium. 
When the antenne possess twelve joints, all pilose, the 
difference between the earlier and later Malpighian tubules 
has ceased to exist. 
The salivary glands are highly developed in all castes and 
at every stage of growth. There is a single pair, as well as a 
large salivary reservoir, such as Miller has described (Pl. 19, 
fig. 7). There is an unpaired external opening in connection 
with the labrum. 
[The supra-cesophageal ganglia are situated as in Thysanura, 
with the olfactory lobes anterior, and the fungiform bodies 
posterior as in Termes lucifugus (see figs. 27—383c, fung.). 
The latter are relatively well developed when compared with 
those of Embiide (fig. 34) or Thysanura; there are two 
on either side, or four in all, not well separated from 
each other. As in other insects, they are characterised by 
1 [*Jen. Zeitschr.,’ ix, pp. 256, 257, pl. xii, fig. 42.] 
