416 
PROF. B. T. LOWNE ON THE 
sections to mistake fibres running between for fibres entering tbe 
optogenic cells. 
Tbe nerve-papilla, from wbicb tbe optic ganglia, tbe optic nerve, 
and tbe retina are developed, gradually grows outwards towards 
tbe dioptron (Plate XXYII. figs. 5-8, n). It is at first covered 
by a layer of columnar cells, which represent tbe epiblast of tbe 
nerve-centre ; from this layer tbe bacillary layer of tbe retina is 
developed. These cells become converted into the retinal end- 
organs. Tbe mesoblastic spongy tissue is gradually absorbed or 
converted into tracheal and connective elements, which ultimately 
form a thin layer between the retina and the basilar membrane 
of tbe dioptron. 
The retina, even when tbe insect is nearly ready to escape from 
tbe pupa, is still separated from tbe dioptron by a space filled 
with branching cells (Plate XXYII. fig. 8, me) and secondary yelk, 
so that tbe supposed entrance of nerve-fibres into tbe dioptron 
cannot be explained by any known process of development. 
The continuity of tbe tracheae of tbe dioptron and those of tbe 
mesoblast is tbe result of the penetration of the bitter between 
tbe great rods during their inward growth ; but during this period 
the nervous papilla is separated by a wide space filled with 
secondary yelk and reticular mesoblast from tbe ingrowing epi- 
thelial structures of tbe dioptron. 
Thus, if my observations are correct, the retina, like that of a 
vertebrate, is entirely formed as an outgrowth from tbe central 
nervous system, while tbe dioptron, like tbe crystalline lens 
and tbe refractive structures generally, is formed from the ex- 
ternal epiblast, wbicb is more or less invaded by mesoblastic 
elements. With regard to tbe retina itself, it is undoubtedly, 
like tbe nerve-centres, no less epiblastic in the insect than in tbe 
vertebrate, as tbe hemispheres themselves, as well as tbe ventral 
ganglia, are formed from tbe embryonic epiblast. 
In conclusion, I would add that it is scarcely fair to expect me 
to prove a negative, i. e. that no nerve-fibres pass to tbe dioptron. 
The onus rather lies with my opponents to prove that the great 
optic nerve does enter tbe dioptron, and to find its terminals. 
Even tbe most cursory glance at tbe works of Dr. Hickson, M. 
Berger, M. Viallanes, and others will show that they have given 
totally dissimilar representations ; of these Dr. Hickson’s are 
correct enough as representations of tracheal and mesoblastic 
skeletal tissues. I would ask, Wbicb of the various structures 
