357 
small scar or indentation in the centre of each facet, like those de- 
scribed by me in Penaeus. 
In both, Belostoma and Tabanus, a corneagen is present, but 
it is evidently in a very primitive condition; there is a large, circular 
opening between every two cells, for the passage outward of the pseudo- 
‘cones, and each pair of cells forms the wall of a deep cup closed 
by the facet only (figs. 1 and 2, c. g.). As this condition is essentially 
a repetition of the early stages of Vespa, we have a satisfactory de- 
monstration that the pseudocone type is the most primi- 
tive. This fact, together with the hair-like nature of the pseudo- 
cones, the resemblance of isolated hair cells to retinophorae or even 
to ommatidia, and the presence of primitive corneal spines in the eyes 
of Vespa, render it almost certain that the pseudocone omma- 
tidium is nothing but a hair-bearing sense bud. In accordance with 
this view, the pseudocone may be regarded as one or more hairs 
Standing in a pit, the mouth of which is closed by the expanded ends 
of the hairs, now converted into corneal facets. The retinula cells 
would there correspond to the so called ,,ganglion cells“ which at one 
time probably extended, as fine processes, into the pseudocone. Indeed 
this condition seems to be retained in some of the Diptera described 
by Craccro (Fig. 4 B.). 
Fig. 4. A. Diagram of an ‘‘olfactory” hair-pit. Partly from Lubbock after Hauser. 
&. Hypothetical primitive ommatidium, a combination of the young stage in Vespa 
whit the adult stage in Tabanus. C. Hair cell from the neighborhood of the eyes in 
Vespa; partly diagrammatic. D. Diagram to illustrate conversion of retinophorae into 
ganglion cells. Z. A hypothetical ommatidium like sense bud. 
