AND ANATOMY OF CERTAIN HYMENOPTERA PARASITICA. 401 



be examined, it will be seen that there are seven stigmata, the first on the 

 fifth segment of the body (text-fig 1, p. 393), the others on the following 

 six segments. The vesicle is still quite clearly marked though much 

 shrunken ; the number of segments is difficult to ascertain with certainty in 

 such specimens. If larvae are boiled in 10 per cent, potash solution till onlv 

 the chitin is left, most insiructive preparations are procured : in text-fio-. 2 

 there is drawn the skin of the adult larva ; segments are clearly marked by 

 the pale line wlieie the chitin thins out to form the joint ; in such prepara- 

 tions there are to be seen fourteen segments altogether where the chitin 

 alternately thickens and then becomes thin. The first stigma is on segment 

 five, the last on segment eleven. The posterior abdominal region is instruc- 

 tive also ; we see a clear thirteenth segment, then a pale joint, then a very 

 narrow but perfectly demonstrable segment, and finally a pale bulb — the 

 terminal vesicle. Now it will be seen that each segmental ring has a row of 

 chitinous areas or teeth, situated behind the stigmn, if this is present in the 

 segment. The first segment has no row of teeth (text-fig. 2), the second 

 ajid succeeding ones, even to the small fourteenth, nil have these chitinous 

 papillse or teeth. Inspection of PI. 24. fig. 6 shows that the brain does not 

 altogether lie in segment one; part of the brain lies in segment two- 

 moreover, in segment two there are no imaginal discs. It seems certain that 

 segments one and two both go to form the head, segment three is the 

 prothoracic, segment four the mesothoracic, segment five the metathoracic. 

 The imaginal discs of the wings (W) are in the meso- and meta-thorax, those 

 of the legs from the third to the fifth segments. Re-examination of the potash 

 preparation in text-fig. 2 now shows that the first stigma is on the meta- 

 thoracic segment, the following six on first six abdominals. After the 

 thoracic segments there are ten abdominal segments, counting the vesicle as 

 a segment. Seurat (12) makes nine abdominals, but figures a rino- in his 

 drawing on page 65, fig. 11, which is undoubtedly the little seoment 

 fourteen, which I have shown to have proper chitinous papillae, and which is 

 undoubtedly a true segment. 



As for other abdominal segments, it is possible that the " bottle-stopper '' 

 joint drawn in PI. 26. fig. 25, at 9, may be partially or wholly formed of 

 degenerate segments. The evidence of early embryonic stages might be 

 conclusive with regard to this point. 



A Hyperparasite (Mesochorus pallidus, Brisch.) of Microgaster connexus. 



In PI. 24. fig. 6, at HP, inside the body-cavity of the Microgaster larva is 

 anothpr, hyperparasitic larva. In PI. 25. fig. 11 is a hyperparasite at X 

 in transverse section. Cocoons of Microgaster bred by me in the laboratory 

 produced some Microgasters and also another insect, Mesochorus pallidus 

 which is drawn in PI. 25- fig-, 7- The hyperparasitized cocoons can be 



