1038 Additions: AGRIAS. By A. H. Fassl. 



form croesus into the entirely constant form sardanapalus of the lowlands of the Central Amazon District. The 

 row of transitional forms from Paia to the Rio Madeira is at present so complete that a separation of croesus 

 and its similar forms from the northern datidia is no more possible. 



I took the typical form croesus in very fine and large specimens on the Rio Xingu as well as with 



vidcaniis. a somewhat shortened red disc of the hind\^ing on the Rio Tocantins, from where also the figirred form vul- 



canus (113 B a cJ) originates with a blue reflection towards the base and anal part on the hindwing. This is 



already a transitional form to A. claudina (from Rio de Janeuo to Bahia), and still more so is a smaller and 



more insignificant form from Alcobaca, the first of the rapids of the Tocantins, in which the quite transcellular 



red spot I'eceding towards the apex of the hindwing, encompassing yet the cell, extends in fine red lines along 



loM. the veins into the disc of the wing : loki form. nov. On the Rio Xingu I captured in numbers only the typical 



form croesus Stgr., the legitimate (^ of which does not exhibit any blue bordering of the red colouring. On 



the Rio Tapajoz, however, there occiu's already a form with a sardana.'palus-h\\ie anal edging of the red disc 



on the hindwing and sometimes also blue colour in front of the red bow of the forewing which was placed 



by Staudinger as a (J to Riffahet's genuine $ of croesus from Chaves (Island of Marajo). I denominated 



this form of croesus decorated with blue of which I possess beside several (J(^ also the very rare, likewise anal- 



micliaell. wards blue $ from the Tapajoz, according to its discoverer: A. michaeli. 



Cachoeira I, the first of the rapids of the Tapajoz, already has a form with a very small and along 



godmanides. the veins dissolved red spot of the hindwing, A^-hich is situate in a larger bluish -violet spot (godmanides Fassl). 



On the Rio JMaues we find a somewhat smaller, very constant local race Avith carmine magnificent 



spots hued with violet \A-ithout a blue bordering, which is more oval in the hindwing and distally less dentate 



pulcherri- than in croesus: pulcherrima Fassl. 



'""• From the next large tributary of the Amazon, Rio Madeira, finally comes the tjrpical sardanapalus; 



but besides there occiu' specimens poor in colours, in which the blue in the forewing is entirely absent. My 

 collector H. Stbympel, however, also succeeded in capturing here 2 genuine (JcJ of sardanapahis which in 

 the midst of the blue disc of the hindwing exliibits yet the large red disc of croesus. I denominate the magnifi- 

 cent new- form completing an unexpected transition from the east from croesus over michaeli, from the south 

 bclsazar. from godmani to the genuine sardanapalus, as belsazar form. nov. (113 B a (J). 



Among the greatest number of the ^,S of sai'danapalus lying before me from the Central Amazon low- 

 lands from Madeira to Peru, Ai-hich are almost quite constantly coloured, I do not possess one specimen with 

 any trace of red in the hindwing; but instead of it the hitherto unlinoM'n $2 of the genuine sardanapalus 

 are most surprisingly of a quite unexpected variability, to such an extent that often all the analogies of 

 colouring which in the aforementioned eastern forms are bound to certain, far remote localities, occur here 

 in different $ forms of sardanapalus verus at the same place and time. As a typical $ form among the 

 35 $$ lying before me I denote the form with a uni-coloured greyish-black hindwing without any trace of 

 ruhrhnedia- blue; since it had been observed by collectors already before, but not captured, rubrimediana I denominated 

 na. ^^ with red-hued medians of the hindwings; 2-ab. purpurea those in which these lines are condensed to a red 

 hrunhilda ^^i^cal Spot of the hindwing; $-ab. brunhilda those specimens where the red m the hindwing is replaced by 

 a blue disc, sometimes yet with a blue distal bordering of the red bow of the forewing; thus a retrogression 

 suprema. to the colouring of the (J. Finally I denominated ab. suprema (113 Bb $) an extraordinarily variegated $ 

 from Teffe exhibiting in the blue brunhilda-spot of the hindwing yet the red spot of the form purpurea. 

 Two $? in which the apex of the forewing is ferruginous which occurs in no other ^4^ria5-form above known 

 coccinata. to me, I described as $-ab. coccinata. 



As to the Andine forms of sardanapalus, we may supplementarily add after a consultation with H. 

 Lathy, that his form hades Lathy has black hindwings without any blue, so that decyanea, which was later 

 on described by Niepelt (p. 570), is to be cancelled as the synonym of it. Moreover, we must insert yet in 

 inicrmedius. Fetjhstorfee's description of the races of sardanapalus on p. 570 the form intermedius Fassl (from the Eastern 

 Cordilleras of Colombia) which was before described by me; it is the northernmost and most scantily coloured 

 form of sardanapahis at any rate, with a dull upper smface as in Agrias aedon to which, however, it does 

 by no means form a transition. 



^4. narcissus does not range as in the previous work of Frithstorfee between phalcido7i and heicitsonius 

 but directly after aedon to which it is very closely allied and perhaps connected with it by transition in hitherto 

 imexplored districts (Venezuela 1). The type occurring to the north of the Amazon invariably shows a red band 

 almost rectangularly touchmg the costal margin. On the western frontier of its range known hitherto, to 

 the north of Manaos, I discovered to my surprise only $$ with a magnificent ochreous-yellow instead of purple- 

 chrysotae- red band, whilst the ^^ were normal; $-var. chrysotaenia (t. 113 Bb). 



A still greater sm'prise on my Amazon exploration was the discovery of a most magnificent, considerably 

 different race of narcissus in the southern Amazon District — from the Rivers Xingu, Tapajoz and Maues — 

 which in contrast with the verj^ constant northern form varies so much at the same place and time that the 

 extreme:it forms, on their' upper sm-face, do no more resemble in the least the exterior of narcissiis. All the 

 representatives of this new southern race are considerably distinguished from the type by the red bow of the 



Mia. 



