276 MEMOIRS NATIONAL ACADEMY OF SCIENCES. vol. xii, 



On examination of alcoholic specimens of stages IV and V, I find on the second abdominal 

 segment a pair of small tubercles situated in a place exactly corresponding to that of the mid- 

 abdominal legs. Each one is about one-eighth as large as a mid-abdominal leg; about as high 

 as broad at the base, rounded, soft, and giving rise to six or seven black setae of uneven length, 

 but with no definite arrangement; the tubercles are situated in the path of the latero-ventral 

 black line. On the first abdominal segment is a minute soft low flattened tubercle situated in 

 the same relative position, but wanting in stage IV. Are these processes to be compared with 

 the supernumerary legs of Megalopyge? Are they the survivals of primitive abdominal legs ? 



Pupa. — In shape and size closely similar to that of Eacles imperialis. It agrees with the 

 head characters, i. e., the eyes, antennae, legs, and wings, differing only in lacking the short 

 stout spines which beset the head and thorax of Eacles, the body being smoother and some- 

 what polished; the hinder edge of the free segments (abdominal) is smooth, not spinose, as in 

 Eacles. The chief difference consists in the shape of the cremaster, which instead of being 

 narrow, slender, and flattened, is very stout, somewhat conical, much contracted at its base, and 

 with coarse pits, some of which are confluent, forming deep smooth furrows or channels. The 

 cremaster ends in two smooth cylindrical spines. Length 40 mm.; thickness of body 13 mm. 



Variation in the terminal spines of the cremaster. — In one example there is but a single median 

 spine at the end of the cremaster, with no vestige of its mate, the spine being central. 



When one comes to the moth one would hesitate at classifying Brahmsea with the Cerato- 

 campidae or the Saturnians; its general appearance forbids this. The wings of the Asiatic species 

 are not falcate, and the mode of coloration of the Brahmseidae is marked by the extraordinary 

 development of undulating lines or bars, which remind one of the geometrid Scotosia undulata. 

 Evidently its style of protective coloration, the green, black, and gray tints and lines, enable it 

 while resting on the bark of a lichen-covered tree to elude observation. 



The venation ( $ and ? ), however, while presenting some important Ceratocampid char- 

 acters, indicates that Brahmsea should be assigned to an independent family. The first subcostal 

 vein (III!) is very short, arising very near the apex of the wing, far beyond the origin of the third 

 subcostal (III 3 ). The second median (IV 2 ) forms an independent vein, as in Ceratocampidae. 

 The discal cell is very small, and the discocellular veins form a reentering angle. In the hind 

 wing the discal cell is remarkably short, not much longer than wide. The subcostal vein (II), 

 instead of being remote from the subcostal as in all Ceratocampidae, is very close to the subcostal, 

 nearly touching it a little beyond the discocellulars. 



The 6" genital armature is of the Ceratocampid type; thesuranal plate being large, broad, 

 slightly bilobed at the end, while the claspers are very broad, rounded at the end, and not 

 mucronate. 



While B. certhia is regarded as the type of the genus, it, or any of the Asiatic species, scarcely 

 appears to be the stem form. For this we shall have to turn to the West African B. lucina, in 

 which the fore wings are produced toward the apex which is somewhat falcate, and the markings 

 suggest the South American Arsenura; the apical oval dark spots recall those of that genus. 



Although there are certain striking superficial resemblances in the larva and pupa to the 

 Ceratocampidaa, the antennas are, however, as in Bomhyx mori and Lasiocampidae, there being 

 but a single pair of pectinatons to a joint. The larva before the first molt is as in B. mori and 

 Endromis, the flattened tubercles giving rise to several (6 to 10) setae. The head is as in B., 

 mori, the epicranium short and broad, the clypeus sunken; the suranal plate as in B. mori and 

 Lasiocampidse. The Brahmaeidse should be associated in a group for which we propose the 

 name Symbombycina, with the families Bombycidae, Endromidae, Lasiocampidae, Liparidae, and 

 Eupterotidae. 



We have, from our studies, been led to infer that the original home of the Brahmaeidae may 

 have been in Africa south of the Sahara region, and that the Asiatic and southeastern species are 

 derived from African forms. They are certainly rather more modified, the wings shorter and 

 broader, and the markings more specialized than in the African B. lucina. 



