UK. .1 I.N I. NA'I. HIST., 13: 21)01 



growth, and the agreement between then segmentation and that recorded lor the 

 corresponding live instar established. A few rough sketches were made of the 

 developing female genitalia, but other features such as ratios of head to pronotal 

 lengths, and details of the fastigial foveolac, tympanal grooves and wing-lobes before 

 reversal, although changing progressively during nymphal development. pro\ed 

 unhelpful in the present context and were not pursued further. 



Certain aspects of the results obtained in 1969 highlighted a need for further 

 investigation and the exercise was repeated, with some modifications, in 1970. 

 Attention was concentrated on female nymphs, 87 of which were collected and 63 

 retained for at least one moult. Because of the possibly adverse effects of freezing, 

 nymphs were immobilised for measurement by gentle compression from above by the 

 bottom of a plastic Petri dish, on which the glass scale was laid, against a film of thin 

 rubber stretched like a drumhead over a cavity below. Provided that at least two days 

 elapsed before moulted nymphs were thus treated, no harm resulted. As before, hind- 

 femora measurements of nymphal casts, or of the corpses of the few that died, were 

 occasionally used when a gap occurred in the data for live nymphs. More attention 

 was directed to the antennae of living nymphs; counts of the flagellar segments were 

 recorded for at least one stage of almost every individual, and confirmed for a few- 

 casts. 



Sites B and C 



During both 1973 and 1975 the approach of 1970 was applied to a total o\~ 41 

 females from site B and 1 1 from site C. Counts of the antennal segments were 

 confined to the later stages of one apparently aberrant individual from site B in 1973 

 and another from this site in 1975. Some sketches were made of the developing 

 external genitalia, particularly those of the second instar. 



Results 

 Site A, 1969 



The measurements of the hind femora have been plotted as histograms in Figs 1 

 and 2. The data for males fall into four discrete clusters, the last two of which are 

 identified by wing character with stages N-l and N. The progress of nymphal 

 development is indicated by the diagonal relationship between the clusters, 

 individuals represented in any cluster in series b to d having contributed, at one 

 stage earlier, to the cluster one position to the left in the series immediately above. 

 This reveals that, of 18 males taken in the first stage, 12 reached the second. 10 the 

 third and 10 the final fourth nymphal stage. Captive rearing has had little effect on 

 the dimensions, and the vertical correspondence between the clusters indicates that 

 nymphs taken in the field exhibited the same four stages. 



In contrast, the data for females fall into five somewhat diffuse clusters. Rearing 

 histories showed that of the 18 nymphs taken in the first stage, two developed 

 differently from the others, wing-reversal occurring at the second moult and full 

 nymphal development being completed in four stages. The blocks representing these 

 two are shown solid. A third wild nymph with hind-femora 4. 1 mm long showed 

 wing-reversal at the next moult and became adult after a further two, while two 

 others (hind-femora 5.5 and 6.0mm) had reversed wing-buds when captured and 

 reached the adult state after two moults. Antennal segmentation (see below) 

 provided strong evidence that these three were also four-stage types, and the blocks 



