BR. i INI nai HIST., 13: 2001 :<" 



hind-femora only 3.9mm long strongly suggested that this also was ot a tour-stage 

 type. In 1975 a third specimen from site B was again regarded as having four nymphal 

 stages on the strength of hind-femora measurements and sequence of wing 

 development; for this the antennal segmentation provided supporting evidence. The 

 blocks representing these three individuals are shaded more strongly. 



For the second instar of most females the ovipositor valves of sternite 8 were 

 present as a pair of blunt triangles, varying in size, but with the apices noticeabh 

 short of the rear end of sternite 9. In this they contrasted with female C. parallelus l< >r 

 which, in the second of the four-stage sequence typical of this species, the valves of 

 sternite 8 normally reach, or even overlap, the rear end of sternite 9. Two of the 

 nymphs apparently approached this conformation, but. for these, sternite 9 either 

 had a strongly excised rear margin or, more probably, bore a curved pre-apical 

 groove that could be mistaken for this; these two shared the common five-stage 

 sequence. For the only nymph with a four-stage history to be thus examined in the 

 second stage this feature did not appear to be markedly different from that of most 

 of the others. It seems possible that the relative positions of these parts may change 

 with the distension of the segments as the stage progresses and that a more extensive 

 appraisal is needed to detect any difference, if such exists, between the two nymphal 

 types. 



Supplementary, 1976 



The four-stage female taken in the second stage from site B and reared to an adult 

 in 1975 remained in isolation until mated with a wild adult male from the same site. 

 From the three batches of eggs laid, 17 nymphs hatched the following year. Two died 

 at the first moult, but the remainder, housed two to a 2-lb jam-jar. were reared as 

 described for 1969. A male and a female, both with reversed wing-rudiments, died in 

 the third stage; 13 others reached the adult state. Of the 15 that could be categorised, 

 five were five-stage and four were four-stage females, six were four-stage males. The 

 antennal segmentation was recorded for the adults and for all female nymphal casts 

 (see Tables 1 and 2), that for the four-stage females agreeing with the findings for the 

 two sharing a known four-stage history in 1969. 



Antennal segmentation 



The antennae of first-stage nymphs consisted, without exception, of a scape and 

 pedicel followed by an 1 1-segmented flagellum. Occasionally, towards the end of the 

 stage, faint indications of future subdivisions became visible through the integument. 

 but, after moulting, the flagellar sheath retained its 1 1 segments. In the second stage 

 the original segmentation was still clearly defined by bold sutures, but slightly weaker 

 new ones were visible on flagellar segments 1 and 6 and. with decreasing intensity . on 

 segments 5, 4 and sometimes 3. Antennal casts of these and succeeding stages 

 normally reproduced exactly the segmentation observed in the living insect. While 

 the antennal characteristics of the first two instars offer an additional feature for 

 identifying these, they provide no clue as to the future developmental pattern. Counts 

 recorded here have been restricted to those for the later instars. when enumeration of 

 the stage coupled with the aspect of the wings is of value in this connection. 



The segmental counts reported by Richards & Waloff (1954) tend to span rather 

 wide ranges, with some overlap between successive stages. Some of this variation 

 may be attributed to the uncertain status of one of the divisions towards the base of 

 the flagellum which, by being often weaker than the others, leaves in doubt whether 



