THE MALE GENITAL TUBE OF SOME OF THE SPECIES OF THE GENUS SCYMNUS (COLEOPTERA, FAM.

The genus Scymnus is one of the largest in the family Coccinellide and one in which the species are very hard to determine on account of their remarkable uniformity in appearance. The adult beetles are small, ranging from 1.1 mm. to approximately 3 mm. in length, rounded to oblong-oval in shape, pubescent, with six ventral abdominal segments showing, and the legs free. The color markings are quite constant, "Color characters have their utility in the separation of species of Scymnus, but some care and no little experience is required in their use" (Horn 1895). Casey (1899) in his revision of the Coccinellidm used color to a great extent in the separation of the species. The prosternal carinm and the abdominal plates or metacoxal lines are other characters which are very useful in this group. The secondary sexual characters of the male are quite distinct in most of the species, but have been mentioned in the description of very few of the species. Leng (1920) lists one hundred and nineteen species from North America, sixty-five of which were described by Casey, seventeen by Leconte, fourteen by Horn, ten by Mulsant, four by Fall, three by Crotch, three by Melsheimer, one by Blatchley, and one by Say. Casey believed that the number of species would be increased in the future instead of decreased. Nothing has been done with the genus Scymnus since Casey’s "Revision of the American Coccinellidm," except the description of four species by Fall and one by Blatchley.


Psyche [October
The genitli of the mle re usually ccepted s being the final criteria for species boutwhose vlidity there is ny doubt.
A study of the genitli will lso show the degree of relationship between species in a group. With these fcts in view the present study was undertaken in the hope that significant variations might be found in the cedgus of species hving only slight external differences.
Since it is very esy to mke erroneous determina.tions in this group, the present pper includes only that mterial which ws compared with the types. All the species were described by Lecont except five, nd the specimens were crefully compared with the Leconte types in the Museum of Comparative Zoology t Cmbridge. The other five species were identified by Leconte, nd were included in his collection. As my identifications gree with his, I felt sfe in using these species in my investigation.
I wish here to thnk Mr. Nthn Bnks for his kind generosity in llowing me the use of the collection in the Museum of Comparative Zoology. I m lso greatly indebted to Mr. C. A. Frost who loaned me his collection with permission to dissect ny of the specimens which I might see fit to use for such purposes.
In previous pper (Wilson 1926) the structure of the Coccinellid cedgus ws described in detail. The general structure of the cedgus of Scymnus is similar to that of other Coccinellids. The bsl lobe surrounds the median lobe nd is very vrible in shape. In such forms s Scymnus americanus Muls., the bsl lobe is quite similar to that of Brachyacantha ursina Fb., which is one of the simplest of the forms described in that pper.
The mteril here presented seems to fll nturlly into four groups, nd if study of the whole genus could be nde it would probably be possible to separate the genus into several groups, ech group showing some relation to the other groups.
The edgus of quadritceniatus Lec. is prticulrly interesting because the proximal end of the median lobe is not flttened nd expnded s in the other species. The median lobe is lso less chitinized and is looped around the median strut. The lateral lobes are large as compared with the basal lobe.
In the second group which consists of americanus Muls., punctatus Mels., cervicalis Muls., and terminatus Say, the proximal end of the median lobe shows specific variations. In fact the proximal end of the median lobe, and the basal lobe, re the structures which are most variable in all the groups.
The oedagus of nanus Lec., semiruber Lec., cinctus Lec., and hcemorrhous Lec., are larger than the cedagi of the second group.
All of the drawings were made at a magnification of ,bout 115 diameters, except in the case of americanus and puncticollis, which were magnified about 85 diameters. The proximal end of the median lobe in nanus nd semiruber are somewhat alike, and resemble that of terminatus to a certain extent. The same area in cinctus and hcemorrhous has in addition a small area, in the upper portion of which are found alternate bands of thick and thin chitin. The basal lobe is also much modified in cinctus and hcemorrhous. It extends through the basal piece, and beyond it anteriorily, for ashort distance. Possibly if more material were available, there would be two groups instead of one, nanus and semiruber belonging to one, and cinctus and hcemorrhous to the other.
The basal lobe of puncticollis Lec., consobrinus Lec., and tenebrosus Muls., is much expnded, and very much larger than the lateral lobes. The shape of the basal lobe is also rather complicated in structure and projects beyond the basal piece anteriorily. At the distal end of the median lobe there are several stiff, bristle-like proiections which point forward. There are also specific differences in the proximal end of the median lobe.
If we consider the genitalia of consobrinus and tenebrosus it ould appear that these species are closely related, while in the key they fll into separate groups, and their relation would never be suspected. The genitalia of the Coccinellidee are rarely extruded and are rather difficult to dissect out, for this reason they will never be of any value as key characters, as in the case of other groups. However, further study of the genitalia of the