DEPARTMENT OF REVISION |
| DIRECTOR Sophronius Marchand |
ABOUT
REVISION
OF THB
LARGE, STYLATED, FOSSOEIAL CRICKETS.
In this paper I have brought under review all the palmated Gryllides known to me in nature or by description, with the exception of Cylindrodes and the minuter forms. My attention was first drawn to them by observing how distinctly they grouped themselves into two divisions, according as the scooping palm was armed with two or with four fingers on the tibia. This seemed to indicate a generic line of demarcation, and the following pages will show the truth of the supposition. I afterwards found that the group had been greatly neglected, and, with the exception of the common European mole cricket, no one species had been well described. Some of the most distinctive features separating the species, such as the size and position of the ocelli, the form of the trochanter, and, in the didactylate species, the form of the sexual abdominal tooth, have been either entirely neglected or but superficially used by entomologists who have written about these insects. There arc other accessible characters not used in this memoir, such as the form of the sternum of the terminal abdominal segments in Gryllotalpa, and the neuration of the tegmina and form of the basal antennal joints in both genera; these ought to be brought forward in a monograph of the groups. The proportions of many parts have been omitted in my descriptions because they could be easily inferred from the very careful measurements which are given; all measurements are made in millimetres; the shortest possible dimensions of the hind femur have been given, and not its extreme length inclusive of the apical prolongations. The longer measurements, especially of flexible, and, therefore, often distorted parts, such as the wings and the anal cerci, are not always quite so carefully made as the others, since conjecture had occasionally to play its part; the measurements from alcoholic specimens arc always exact, and, in any case, the error is but trifling, and the multiplicity of the measurements such as to prevent any general misstatement.