84 GLOSSOTROPHIA. By L. B. Prottt. 



romanaria. G. romanaria Mill. (3 1) is apparently a variable species, and has been much confused with the allied 



species, in particular rufomixtata. The scaling of the dark patches is arranged as in that species, but the fluted 

 appearance even more pronounced. That it cannot, however, be a form of rufomixtata is proved by the 

 tongue and by the early stages, as has been pointed ont by Rebel. The tongue, although long, has not nearly 

 the abnormal length which it attains in that species. I have had no material in undoubted romanaria for mea- 

 suring, and it is impossible to estimate the length when it is rolled up; but in an example of the Syrian sub- 

 species (or close ally) semitata of which I relaxed and extended the tongue I found it 11 — 12 mm long. It 

 looks less in the type-form, and indeed must be so if Milliere's figure of the pupa is correct; for that 

 figure shows no elongate tongue-sheath. The tongue of confinaria, it may be added, measures about 18 mm, 

 but freshly killed examples are needed for accurate work at this question. G. romanaria is generally of small 

 size, in its typical forms rather glossy, the dark dusting not being strong enough to give the wings a more rough- 

 ened appearance. The shades of colour are soft and delicate but sometimes rather bright, as there is often 

 a strong tinge of pinkish or light red; the costal spots and the lines are brown, sometimes with a decided 

 olivaceous tinge ; the shades before the subterminal line greyer. The pale apical patch of the forewing is usually 

 rather conspicuous, except in the lighest forms. The postmedian Ime is not appreciably toothed near the 

 costal margin. Underside almost entirely without markings. In the <$ antenna and hindleg I find nothing 

 distinctive from confinaria or rufomixtata. The larva feeds on the leaves of Linaria and Antirrhinum. It 

 resembles that of Acidalia marginepunctata. Elongate, cylindrical, without lateral flange. Head small. Body 

 fleshy grey, sometimes more clay-coloured, ventrally whitish; dorsal line fine, double, brown; tubercles and 

 stigmata black. The moth is double-brooded, appearing in the sprmg and in July. True romanaria (which 

 according to Milliere varies Kttle) occurs in Central and Southern Italy, Sicily and Tunis; probably also in 

 Spain and Algeria, but there is still much work to be done in arranging and classifying the closely allied 

 pMlipparia. forms. I only indicate one or two which I have been able to examine. — philipparia suh.s-p. nov. is a much 

 darker form from Philippe ville, Algeria. Reddish brown, about as in our figure of falsaria (4 h) the lines and 

 subterminal shades darker brown, all showing very strongly, with the lens, the fine transverse striation. The 

 pale apical spot on forewing and a pale line at base of fruage on both mngs, though not really lighter than 

 in typical romanaria, are more conspicuous on account of the darkening of the ground-colour. (5^? in the 

 PiJNGELER collection, bred ab ovo by K. Andreas in September, 1910. The larvae fed on Caryophyllaceae. 

 I have a larger and lighter 9 (the dark lines and cell-spots consequently showing up more distinctly) from the 

 same locality, taken m May 1909, which is still a good deal darker than typical romanaria. It is however, pos- 

 sible that larger material will show philipparia to be merely an extreme aberration. It is perhaps the N. Afri- 

 can form indicated bySTAUDiNGER under rufomixtata as a transitional form to diffinwia; but it has not the 

 semHafa. coarse dark speckling of the former, and differs from both in the shorter tongue. — semitata subsp. nov. 

 represents romanaria in Syria. It was formerly recorded by Pungeler as a small light form of rufomixtata, 

 but he now determines it (in litt.) as "romanariaV\ I suspect it will prove to be a species distinct from 

 both, and indeed intermediate. The tongue seems to be longer than in typical romanaria. The antennal cilia- 

 tion appears appreciably shorter, though similarly arranged. The ground-colour is of a very pale sandy hue, 

 slightly tinged with reddish, very similar in colouring to coenosaria from the same locaUty; in some specimens 

 the coloiiring is a little deeper, in others a little paler, but it does not seem very variable. It looks less glossy 

 than typical romanaria, the wings being more strongly dusted with dark scales; but these are reddish brown, 

 not so blackish nor so coarse as in rufomixtata. The costal spots of the forewing are as a rule strongly developed, 

 that at the origin of the postmedian line black, or almost black. The average size is at least as small as in 

 romanaria, perhaps slightly smaller. The type specimen, from Baalbek, is in my collection, and was taken by 

 Mr. P. P. Graves at light in May 1905. I have also a second ^ with the same data, and a series of both 

 sexes from "Syria" without moi'e exact locality, Herr Pungeler has others and there is one in the British 

 Museum from the Leech collection. 



isabdlaria. G. isabellaria if '/7/. (7 c) is not certainly known to me, but I have no doubt it belongs in the immediate 



vicinity of romanaria. To judge from some specimens similar in coloration to Milliere's figures I should 

 be inclined to suppose that it was only a richly-coloured Spanish form of that species. But it must not 

 be forgotten that Milliere, who knew the larvae, regarded the two as distinct species and it is possible that 

 isabellaria is a scarce species awaiting rediscovery. The form from Spam and Portugal which I provisionally 

 refer here shows one slight structural difference which should be at least subspecific; the (J antennal joints 

 seem to project still more strongly. The ground-colour is reddish, the markings brown with a rather strong 

 shade of olive, the dark dusting stronger than in romanaria but scarcely so strong as in rufomixtata; the striated 

 scaling of the dark outer shade as in those species. The pale subterminal liiae is conspicuous and thickens 

 considerably between the radials and near the anal angle, about as in Acidalia marginepunctata; the distal 

 shading beyond the subterminal line is so slight that the subterminal itself almost appears to be extended 

 as far as to the black terminal marks. The dark lines are rather thick and not very sharply exjsressed. 

 The wings seem sometimes rather broader. Milliere in erecting his isabellaria speaks of the "feebly pectinate 



