THE SOUTH AFRICAN ODONATA. 589 



As a rule, our forms are found as far north as the KiHiuan- 

 jaro district. There are no distinctively South African species 

 of Odonata, but several South African varieties. Orthetnim 

 caffrum, for example, first described from Durban, extends from 

 Kihmanjaro to Capetown. In the " Selys Catalogue," over two- 

 thirds of our species in all genera are listed as coming also from 

 Kilimanjaro. 



'he correspondence with the Indian insect fauna as to genera, 

 but not as to species, which is noted in other orders, is found 

 among the Odonata. Twenty of our genera, out of a total of 

 about thirty, have Indian representatives. Only one species, so 

 far as I have been able to find, is common to our Odonata fauna 

 and that of India, namely, Pantala flavescens. 



Rather remarkable is the occurrence of all the subfamilies 

 of Odonata save one — the Cordulegasterimc — in our South 

 African fauna. (In the Indian fauna even this is found.) 



The subfamilies Gontphinoe and Cordulitkc, dominantly 

 northern, are each represented by a few species. Remarkably 

 enough, the Goniphince are represented by many individuals, and 

 the Cordulincc bv very few. This I attribute to the adaptation of 

 Gomphine nymphs for burrowing in silt in rivers. 



T wish to conclude by recording a protest against the whole- 

 sale elevation of varieties to species-rank, and of intergrading 

 variations to varietal-rank, which has been done by overseas 

 workers on South African Odonata. Orthetritm stemmalc, for 

 example, divides into several varieties in South Africa ; but as 

 I have in my collection different " varieties," taken in the same 

 spot on the same stream on the same day, I am loath to accept 

 them as separate forms. In the common red Trithcmis forms, 



inulata, etc., there is much variation and intergra'ling. which 

 render the distinctions in the published descriptions impossible to 

 follow. It appears often as if the mere fact of a pair of specimens 

 being different in one cross-vein, and coming, one from Delagoa 

 Bay and another from Natal, had been deemed cause for erecting 

 each into a species. Against such " splitting," based on very small 

 series, I wish most emphatically to protest. My protest may, 

 perhaps have more value when I add that, in my own collection, I 

 have complete intergrading series for Trithemis anmdata and T. 

 artcriosa, as well as for the various forms of Orthetrmn- sfeiiunale. 



CROSSING THE NORTH AFRICAN AND SOUTH 

 AFRICAN OSTRICH. 



By Prof. J. K. Dui-kden. M.Sc, Ph.D., A.R.C.S. 



(Not printed.) 



