The primitive insect, in acquiring the power of flight, must initially have used the existing thoracic musculature to produce the deformation of the thoracic wall required to move the wings: the dorso-ventral muscles inserted on the leg base were, even without change of attachments, used in the capacity of indirect wing levators, for by contracting in mass they depressed the tergal wall and so raised the wing; the pleural (epipleural) muscles, by pulling on the wing, became 'direct' wing depressors; subsequently the development of phragmas enabled the longitudinal tergal muscles to function as weak indirect wing depressors, and to become, eventually, even the principal wing depressors. Such, at least, is the evidence derived from Orthoptera. In higher insects (but not Lepidoptera) evolution of the flying mechanism entailed the development of an increasingly rapid wing beat, leading in Hymenoptera and Diptera to very high-frequency wing vibration with a minimum of thoracic deformation; in the wing musculature it led to the evolution of a completely novel type of muscle in which high-frequency unfused contraction, under almost isometric conditions, replaced low-frequency tetanic shortening. A thoracic muscle, functioning in its original capacity as leg muscle, can only to a restricted degree accept the additional role of wing vibrator, and must have become increasingly useless in its initial role as it developed its new function. In Homoptera, which have been examined in some detail, such a crisis seems to have occurred: the cicadas and various fulgorids have to a surprising degree diverted muscles to the exclusive role of wing vibrators; others, like jassids, have distributed the two functions more evenly in different muscles, and there are signs (in Cercopidae) that even new muscles are generated from existing muscles to act as wing vibrators. A surprising feature in some Homoptera is the co-opting of certain abdominal muscles into the flying musculature, to meet the deficiency of metathoracic muscles. In the Diptera, the only other order examined, the wing vibrators are reduced to a small number; most of the dorso-ventral muscles associated with the leg base disappear, and the tergal wall of the mesothorax is an almost exclusive, in some forms even exclusive, alinotum. Here departure from the primitive orthopteran condition is at its greatest. There is, on the other hand, a strong development of new steering muscles in Diptera. Change in physiological property of the muscle tissue is attended by changes in the histology of its fibres. The orthodox view that the wing musculature of higher insects is composed of fibres strictly comparable with those of other muscle is upheld, but only after a re-assessment of the latter. It appears that in normal arthropod fibre the fibrils are actually composed of a small number of contractile myofibrils bound into a unit by 'ground substance' in which they are embedded; for this compound fibril the term 'sarcostyle' is adopted in this new sense. In the higher, and even intermediate groups, in which the tendency is toward high-frequency isometric contraction, the sarcostyles undergo remarkable thickening; these are the coarse fibrils first described by von Siebold (1848) and now well known for all the higher orders where the wing beat is rapid. In the sarcoplasm the most noteworthy change is the disappearance of the Cohnheim reticulum even within Orthoptera, and in the higher groups a liquifying of the sarcoplasm, presumably to promote mobility of the tissue; it is to the liquid sarcoplasm that the tissue owes its unique tendency to dissociate into fibrils. With these changes is associated an increased development of sarcosomes, which enlarge and adopt a direct relation to the cross-striation. Of the cross-membranes a Z-membrane completely transecting the fibre is an integral component of normal striated muscle fibre; a similar M-membrane is certainly present in vertebrate fibres, but (through lack of specific technique?) can only in exceptional cases be seen in arthropod muscle. Both membranes have appeared in the wing-muscle fibres, even in coarsely fibrillated (Siebold) muscle. It seems to be the cross-membranes, usually helicoidally disposed, that impose on the fibril bundle its pattern of cross-striation. The last criterion for assessing the presence of a muscle fibre is a sarcolemma; in all cases examined such a membrane has appeared, even in species in which its presence is currently denied. In most blattids the wing-muscle fibres do not show any recognizable adaptation to flight; in other primitive Orthoptera the Cohnheim reticulum is retained, and the only perceptible adaptation to flight is an inconstant increase of sarcosome content, and the penetration of a few blindly ending tracheae into the fibres. Yet in Acridiidae a rich sarcosome content has developed, and there is an abundance of intracellular tracheae leading even to closed net formation. In Homoptera, where flight is more highly developed, the musculature seems to be transitional to that of higher insects; the sarcostyles commonly confer a special pattern on the cross-cut fibre, and may be thin or of the thick (Siebold) type, but always respond well to electric stimulation; sarcosomes are abundant and the tracheae form closed nets to a degree far beyond those found in Orthoptera. In the Diptera, the only higher order examined, shortening of sarcostyles in response to electric stimulation is restricted; sarcosomes are always abundant, and the tracheae always form closed nets within the fibre. Different families are distinguished by the fibre pattern of their wing muscles. In some the fibres are numerous; in most they are few and generally constant in number, and wherever the insect is large this leads to giant fibre formation. Development of giant fibres is attended in Diptera by elaboration of their internal tracheae beyond anything found in other insects; usually there is a development of a highly elaborated pattern sometimes even having a relation to the cross-striation. In Homoptera, as in Orthoptera, the motor-nerve fibres end by Doyere end-organs just under the sarcolemma; in various Diptera the innervation has been found to be by numerous nerve twigs all along the fibre, ending, not just under the sarcolemma, but deeply within the interior of the fibre. Throughout the Orthoptera enlargement of the flight muscles is attended by fibre proliferation; specific flying muscles develop by repeated division of rudimentary muscle fibres; muscles that function in double capacity enlarge with growth of the nymph, the most intense fibre cleavage being apparently without effect on the functioning of the tissue. Only in the last instar do the fibres acquire the special histological characters that adapt them to flight. In Homoptera, also, there is evidence of fibre cleavage. In cicadas the entire wing musculature is generated by the repeated division of a few rudimentary fibres. In various other Homoptera, where the wing musculature is composed of many fibres, there is cleavage either of functional nymphal fibres (cercopids) or of rudimentary fibres arising from myoblasts in the minute nymph; only in jassids with small fibre number is cleavage suppressed. But in all these forms (except cicadas) a new histogenetic process soon supervenes, formerly free myoblasts becoming incorporated in a remarkable manner into the young fibres; the myoblasts prolong, either singly or in chains, alongside the enlarging fibres, successively adding to the latter a new fibril with investing sarcoplasm, and with attendant nuclei, and these new fibrils may or may not be subject to fibril cleavage. Among the Homoptera the most specialized process is that found in small jassids, where each myoblast contributes to the fibre a single fibril. In Homoptera the young fibres are usually, from the beginning, faintly cross-striated, and to the existing pattern the newly added fibrils accommodate themselves. In Diptera the myogenesis is concentrated in the pupal period, in contrast to that of Homoptera, where it occupies most of the nymphal period. It is a process in which multitudes of myoblasts co-operate, each apparently supplying a single fibril, as in jassids; but cross-striation is here deferred till the full fibril number has appeared. The meaning of this type of myogenesis is unknown; but it emphasizes the fact that the formation of a striated muscle fibre is a more plastic process than we have hitherto suspected.