446 PERCY SLADEN TRUST EXPEDITION 



conspecific with the Madagascar Hemicordulia similis. Teinohasis alluaudi is another 

 species which is common to Madagascar and the Seychelles. The forms which are really- 

 peculiar to the Archipelago are Orthetrum stemmale ivrighti, Pseudomacromia luctifera, 

 Gynacantha sty lata, Allohstes maclachlani, Allolestes nigra, and Leptocnemis bilmeata. 

 Pantala Jiavescens belongs to no particular zoogeographical region. The remaining species 

 appertain to the Oriental fauna, and are as follows : — Diplacodes trivialis, Zyxomma petio- 

 latum, Tholymis tillavga, Anax guttatus, and Agriocnemis pygmcBa. An analysis of the 

 indigenous species and subspecies reveals the interesting fact that the three Anisoptera — 

 Orthetrum stemmale ivrighti, Pseudomacromia luctifera, and Gynacantha stylata — are 

 allied to forms found in Africa or Madagascar, while the three Zygoptera — Allolestes 

 maclachlani, A. nigra, and Leptocnemis hiliyieata — have affinities with Asiatic forms. 



It may be taken that the Odonate-fauna of the Seychelles is now fairly well known, 

 and the apparent deficiencies in it are quite remarkable. For instance, notwithstanding 

 decided relation with the Oriental region, there is no representative of the characteristic 

 Asiatic genus Neurothemis (Libellulinse). The Libelluline genera Palpopleura, Croco- 

 thems, Brachythemis, ai:id Trithemis are also unrepresented, although they occur quite 

 commonly both in Asia and Africa. Furthermore, it is worthy of remark that the 

 important Subfamilies Calopteryginse and Gomphinse have never been met with. 



The presence of larval water-mites (Hydrachnidse) is a very noticeable feature of 

 some of the specimens of Tramea limbata from the Seychelles, and also of the Tramea 

 hasilaris from Aldabra. In some instances the thorax carries only two or three of the 

 large ruddy mites, while in other cases the sides and under surface of the thorax are 

 thickly encrusted with them. It would appear that these Arachnids are not met with 

 at all commonly upon the bodies of Anisopterous Dragonflies, and one of the few other 

 cases of the kind which have come under my notice concerns an Australian Tram,ea. 

 Sympetrmn meridionale Selys furnishes a notorious instance of larval water-mites 

 clustering in great numbers on the wings of an Anisopterous species, and Mr K. J. 

 Morton possesses a specimen from Asia Minor whose wings bear not less than 220 of 

 these curious little creatures. Among the Zygoptera from the Seychelles, at least one 

 specimen of Agriocnemis piygmcBa has mites on the sternum. The nature of the relation 

 subsisting between Acari and their Dragonfly hosts is not at present well understood. 

 (See on this subject Berlese, Gli Insetti, vol. ii. pp. 13 and 26.) 



POSTSCRIPT. 



Since the foregoing was written (June, 1912), I have seen several recent papers by Dr Ris bearing more or 

 less directly upon the species discussed, but unfortunately it has not been found possible to insert references to them. 



One of these publications is fasc. xiv. of the Selysian Catalogue, which includes Dr Ris' revision of Pseudo- 

 macromia. The Seychellean Dragonfly originally described as Zygonyx luctifera, and formerly believed to have 

 affinities with certain Asiatic foi'ius, is now definitely referred to the African genus Pseudomacromia, as fore- 

 shadowed in the present paper. 



Again, the cj anal appendages of the species designated herein as Lestes ochraceus (forma) are very similar 

 to those figured by Dr Ris for Lestes unicolor, M'Lach., from the River Nile (Gondokoro) and from Madagascar 

 (Sitzungs. Akad. Wissen. Wien., Mathem.-naturw., cxxi., p. 153 (1912)). In connection with the preparation 

 of my report I compared tiie typt; ^ of L. U7iicolor with other males from the Percy Sladen Trust Collection, 

 and I made a note at the time of this similarity in the matter of the anal appendages, in addition to other 

 points of agreement between the insects compared. I also found a still closer agreement between the type 

 $ of L. unicolor and the ? from Cosmoledo. Nevertheless, in view of certain diiferences which I discovered 

 and to which I attached some importance, I felt disinclined to identify the series from Aldabra and Cosmoledo 

 with M'Lachlan's types from Madagascar. 



24 January, 1913. H. C. 



