A SUPPLEMENT TO THE WORLD REVISION OF ODONTOMACHUS ( HYMENOPTERA : FORMICIDAE ) ' By

In a recent publication (Brown, 1976, Studia Entomologica 19: 67-171) I reviewed the world fauna of the ponerine ant genus Odontomachus. While that contribution was in press, and since its appearance, some significant new Odontomachus material, and information about material, has come to my notice. It seems appropriate to supplement the revision now, while it is fresh, by offering the new information for incorporation. The first and most important addition is the description, of a new (twentieth) species in the haematodus group.


Psyche
[September-December distinct anteriad on each segment, while the costulae predominate posteriad; this is especially true in the eastern (Surinam and Guyana) samples, in which the costulae become virtually obsolete on the anterior half of each tergum.The sculptured surface is opaque to sericeous-opaque, depending on magnification and lighting.The fourth gastric tergum, largely retracted, appears to be finely and superficially punctulate and subopaque.Underside of gaster shin- ing, finely reticulate.
Other minor average differences from O. bauri are as follows: Costulae of the head and trunk a trifle more strongly shining (about 6-7 costulae per 0.1 mm square on each side of the vertexal midline).Striation of petiolar node finer than in most bauri, oblique.The node itself is a little less "dome-shaped", less convex in outline behind, and therefore more as in O. haematodus, but the metasternal process is low, and not produced into paired sharp teeth, as it is in O. haematodus.
Apical spine of petiole straight and only weakly back-tilted.
Pilosity and pubescence copious and distributed much as in O.
bauri, but on the average a trifle shorter and less dense; 4-8 stand- ing hairs on pronotum; standing hairs of body about 0.2-0.45mm long.Mesopleura obliquely striate, with a few coarse punctures in the mid-section; in the eastern samples, the mesopleural sculpture is effaced, smooth and shining in the mid-section, and the punc- tures here are smaller.Color black in Ecuadorean samples, with dark brown legs; antennal funicuU and tarsi yellowish-brown.
Eastern samples dark reddish-brown to piceous (some possibly faded), funiculi and tarsi yellowish-brown.Palpi segmented 4,3.In some specimens, the longitudinal striation of the metanotum even continues for a short distance onto the anterior part of the propodeum.
Queen and male unknown.
One paratype worker from southern Guyana: Oronoque River, 2°4rN, 25 July 1936, by N. A. Weber, no. 594.Paratypes in MCZ and U.S. National Museum of Natural History, Washington, D.C.     1977]   Brown -Odontomachus 283 During work on the revision of Odontomachus (Brown, 1976, Stud.Ent.19:67-171) I had in hand a single specimen of this new species, the paratype from Guyana listed above.The specimen lacked both antennae, and in a complex as difficult as the haema- todus group, it was unwise to add yet another sibling species on the basis of a sample that might have been only a freak variant of O. bauri.Since the revision went to press, however, I have been able to study the 4 additional specimens from two widely separated hylaean locahties, which convince me of the high probabihty that this phenon represents a distinct species.
In the key to neotropical Odontomachus species (Brown, 1976,   op.cit

Odontomachus laticeps
In my discussion of this variable species in 1976 (p. 154 ff.),I found no South American records certainly attributable to it, but Dr. D. R. Smith has now called my attention to a small number of worker specimens in the U.S. National Museum from Bolivia that probably do belong to O. laticeps.These samples come from Covendo, Huachi and Cachuela Esperanza, all in Beni Province, and all were collected by W. M. Mann.These workers are brown in color, with legs usually lighter, and they all have the gastric dorsum predominantly finely but distinctly longitudinally striate and opaque; the bottoms of the striae are punctulate.The size is a httle smaller than in Central American laticeps (HL 2.48, HW 1.67 mm in an average specimen from Huachi, Bolivia), but the cephalic index ( 67) is within the range of Central American laticeps, which have CI 66-73.
The main difference comes in the shape of the apical spine of the petiole: This is lower and thicker in side view in the Bolivian sam- ples, but still sharply back-tilted: figure 14 (p. 156)in the 1976 re- review shows the slender spine of an average Central American specimen.The Bolivian population{s) could represent another sib- ling species of the haematodus group restricted to Bolivia (or even to the drainage of the Rio Beni), but the close relationship to O. laticeps is obvious, and I prefer to assign them to that species until we know more about the Bolivian Odontomachus in general.All previous records of O. laticeps from South America are based on O. biumbonatus or other species.
Odontomachus mormo I described this giant species in the 1976 review (p. 161)from 3 specimens collected in transandean Ecuador.Now Stewart and Jarmila Peck have sent me more specimens, including a partial nest series with queen, from the Rio Palenque Research Station, 47 km SW of Santo Domingo de los Colorados, Pichincha Prov., Ecua- dor.The species remains known only from western Ecuador.

Odontomachus bradleyi
This species, also described in 1976 (p. 133) from Dept.Junin in central Peru, has now been rediscovered by the Pecks in eastern Ecuador: 23 km NE of Baeza, Napo Prov., 4 March 1976, nest under log in pasture.This record is a long extension of range, and shows that the species is probably widely distributed along the east- ern flank of the Andes in Amazonian-drainage valleys.
., p. 112), O. scalptus keys uneasily to couplet 3, first lug (O.yucatecus) -"uneasily" because some O. scalptus samples have striation fairly distinct, even if "mixed", over the anterior half of gastric tergum I. O. scalptus differs primarily from O. yucatecus in having the gastric dorsum distinctly and sharply, if finely, sculp- tured.The only other species with both mesonotum and gastric dorsum longitudinally striate is the sympatric O. caelatus, but this species is larger, has the gastric dorsum regularly striate (not punctulate), with a few, large, blunt hairs, and virtually obsolete pubescence.The appressed and decumbent pubescence of O. scalptus is abun- dant and conspicuous by contrast.Taylor informs me in litt.that his recent researches in northern Queensland indicate that the variant of O. cephalotes having a densely punctulate or striate gastric dorsum and numer- ous standing hairs on the pronotum is actually a separate and dis- tinct species occurring in the area near the base of Cape York and westward to Arnhem Land and beyond, perhaps to the Kimber- leys.He finds it only in savanna or savanna woodland areas, and I recall that a form answering this general description is common nesting in termite hills in northern Queensland.Taylor writes that he has been calling this form O. turneri, but that it might possibly have a prior name (semicircularis Mayr?).I have thought of this form as corresponding to Forel's var.ajax, but since it shows con- siderable geographical variation, the whole situation needs re-study in the light of the new material now gathered in the Australian National Insect Collection. Psyche[September-December