Etymologically (σπασμὁς, σπἁω I draw) the word signifies a twitch, but as it is unfortunately considered a synonym for convulsion, the two expressions are used indifferently in medical parlance, though the desirability of restricting the application of the former has more than once been indicated. Littré's definition--"an involuntary contraction of muscles, more particularly of those not under voluntary control"--may appear somewhat idle, as the contraction of muscles not under the influence of the will can scarcely be other than involuntary. His intention was, no doubt, to reserve spasm for convulsive phenomena in non-striped muscle fibres; but in this limited sense the term has not met with acceptance, and it remains equivalent to "involuntary muscular contraction," whatever that may mean. Thus interpreted, it is applicable to any and every involuntary muscular movement, physiological and pathological, to the inco-ordination of tabes, to chorea, athetosis, tremor, etc. Rather than imagine a new substantive to characterise certain of these muscle contractions, we may retain the word in a somewhat wider though equally precise sense, and follow the distinction drawn by Brissaud[9] in 1893: "a spasm is the result of sudden transitory irritation of any point in a reflex arc; ... it is a reflex act of purely spinal or bulbo-spinal origin." By definition, then, _a spasm is the motor reaction consequent on stimulation of some point in a reflex spinal or bulbo-spinal arc_. To differentiate between the reflex, which is physiological, and the spasm, which is pathological, we may add as a corollary: _the irritation provocative of the spasm is itself of pathological origin, and no spasm can occur without it_. The anatomo-pathological substratum of a spasm is, then, some focus of irritation on a spinal or bulbo-spinal reflex arc, which may be situated in peripheral end organ, in centripetal path, in medullary centre, or in centrifugal fibre. Whatever be its localisation, it will determine a spasm in our sense of the word. Cortical or subcortical excitation, however, as well as peripheral stimuli, may provoke these bulbar and spinal centres to activity. Irritation of a point on the rolandic cortex, or on the cortico-spinal centripetal paths, is followed by a motor reaction exactly as with afferent impulses; the sole change is in the route taken by the centripetal stimulus; the reflex centre remains bulbo-spinal, and the efferent limb of the arc is as before.