The scorpion never stings _backwards_, but always _in front_ of itself. It delivers stabs with its sting in two distinct ways. For the purpose of defending itself from attack it elevates its abdomen into a bow, and then regains its former position by suddenly unbending it. To strike an animal, such as a spider, which serves it for food, the scorpion seizes it with its pincers and holds it as in a vice. Then it raises its abdomen, brings the end of it close to its captive, and, with a lever-like movement, drives the sting into its body. The victim immediately becomes paralysed and motionless.[115] The poison-glands of a _Scorpio occitanus_ from the South of France contain about 1 to 10 centigrammes of a toxic liquid, capable of furnishing 10 to 15 per cent. of dry extract. This liquid is decidedly acid; it reddens litmus paper and is miscible with water. Its physiological effects are especially intense in the case of the arthropods upon which the scorpion habitually feeds, and in that of vertebrates in general. Batrachians, fishes, birds, and mammals are extremely susceptible to this poison. Half a milligramme of dry extract injected subcutaneously is sufficient to kill a guinea-pig, and 1 milligramme is lethal to the rabbit. In poisoned animals there is first observed a period of violent excitement, accompanied by very acute pains; these are followed by muscular contractions, and finally by paralysis of the respiratory muscles, as in the case of intoxication by cobra-venom. The effects of scorpion-poison, which clearly indicate the presence of a _neurotoxin_, have been very well described by Valentin,[116] Paul Bert[117] and Joyeux-Laffuie. Kyes[118] has prepared a _lecithide_ from scorpion-venom, which hæmolyses red corpuscles as do the lecithides of cobra-venom, and I myself[119] have established the fact that the antivenomous serum of a horse vaccinated against cobra-venom effectively protects mice and guinea-pigs against intoxication by the venom of _Scorpio occitanus_; this has been verified by Metchnikoff. There is, therefore, a close affinity between this venom and that of COLUBRINE snakes. On the other hand, it has been shown by the investigations of C. Nicolle and G. Catouillard that the same antivenomous serum has no effect upon the much weaker venom of the scorpion of Tunis (_Heterometrus maurus_), which, in the case of man and mammals in general, scarcely does more than produce a transient œdema at the point of inoculation.