278 MR. F. BALFOUR-BROWNE ON THE [ Mar. 2, 
moult. This statement is true of Agrion, Ischnura, and Pyrrho- 
soma, and may be true of Hnallagma and Hrythromma, but I 
imagine that in these cases wing-development begins rather later. 
These ridges lie diagonally across the second and third thoracic 
segments on either side of the tergum, and are the result of the 
upward growth of the pleura of the meso- and metathorax. 
The rate of development from the first differs somewhat in 
different individuals. In the stage succeeding that in which the 
ridges first appear there may be a small bud on the inner side of 
each ridge, representing the commencement of the “wing,” or 
this may not appear until a stage later. 
As the buds grow larger the ridges also develop, those of the 
mesothorax gradually spreading anteriorly over the tergum until 
they meet in the middle line. Posteriorly they do not approach 
one another, so that a triangular patch of the tergum is left 
exposed between them. The metathoracic ridges joi on to the 
posterior edges of the mesothoracic ones, and taper away 
posteriorly towards the sides of their segment, the posterior 
margin of which they slightly overlap. Thus a large part of the 
tergite of the metathorax is left exposed. The “ wings” develop 
in such a way that they lie more or less flat on the sides of the 
dorsum in the wedge-shaped depression between the ridges, and 
extend backwards over the anterior abdominal segments. Towards 
the end of the final stage, however, when the thorax of the 
imago is forming within that of the nymph, the “wings” are 
somewhat raised and tilted. 
At each moult an advance in development is visible, but it 
differs very considerably in the earlier stages in different indi- 
viduals.’ For instance, a condition when the anterior ‘“ wings” 
just overlap the posterior edge of the metathorax, and the 
posterior wings just cover the first abdominal segment, might 
represent the 8th, 9th, or 10th stages, apparently depending on 
the general body-growth of the nymph and the number of stages 
through which the nymph is going to pass. There is a distinct 
connection between the later moults and the proportion of 
increase in the length of the wings. For instance, if the anterior 
wings reach in one stage almost. to the posterior margin of the 
metathorax, their increase at the next moult will be about 
33 per cent. of their length. At the next moult the increase 
will be about 40 per cent., while at the next, whether the 10th, 
11th, or 12th, &e., the increase will be about 50 percent. This 
is a large increase, and is, in every case I have observed, evidence 
that the nymph has reached its final stage. The “ wings” now 
cover the first three abdominal segments, and often extend half- 
way over the fourth, and any nymph found with such long wing- 
cases is in its last and generally its longest stage. 
The last nymphal stage is characterized by these long “ wings” 
and by the 7-segmented antenne, and the one or two stages 
previous to that are fairly well-marked by the rapid growth of 
the wings. The first four stages in the life-history are also easily 
recognized by the state of the antenne and of the mask. In my 
