Kobert therefore sought to prepare extracts which would contain the active principles of the poison and which could be injected in definite quantities directly into the blood of the experimental animal. For this purpose various parts of the spiders were rubbed up in a mortar with distilled water, or physiological salt solution, allowed to stand for an hour, filtered, and then carefully washed, by adding water drop by drop for twenty-four hours. The filtrate and the wash-water were then united, well mixed and, if necessary, cleared by centrifuging or by exposure to cold. The mixture was again filtered, measured, and used, in part, for injection and, in part, for the determination of the organic materials.

Such an extract was prepared from the cephalothoraces of eight dried specimens of the Russian Latrodectus and three cubic centimeters of this, containing 4.29 mg. of organic material, were injected into the jugular vein of a cat weighing 2450 grams. The previously very active animal was paralyzed and lay in whatever position it was placed. The sensibility of the skin of the extremities and the rump was so reduced that there was no reaction from cutting or sticking. There quickly followed dyspnœa, convulsions, paralysis of the respiratory muscles and of the heart. In twenty-eight minutes the cat was dead, after having exhibited exactly the symptoms observed in severe cases of poisoning of man from the bite of this spider.

These experiments were continued on cats, dogs, guinea pigs and various other animals. Not only extracts from the cephalothorax, but from other parts of the body, from newly hatched spiders, and from the eggs were used and all showed a similar virulence. Every effort was made to avoid sources of error and the experiments, conducted by such a recognized authority in the field of toxicology, must be accepted as conclusively showing that this spider and, presumably, other species of the genus Latrodectus against which the clinical evidence is quite parallel, possess a poison which paralyzes the heart and central nervous system, with or without preliminary stimulus of the motor center. If the quantity of the poison which comes into direct contact with the blood is large, there may occur hæmolysis and thrombosis of the vessels.

On the other hand, check experiments were carried out, using similar extracts of many common European spiders of the genera Tegenaria, Drassus, Agelena, Eucharia and Argyroneta, as well as the Russian tarantula, Lycosa singoriensis. In no other case was the effect on experimental animals comparable to the Latrodectus extract.

Kobert concludes that in its chemical nature the poison is neither an alkaloid, nor a glycoside, nor an acid, but a toxalbumen, or poisonous enzyme which is very similar to certain other animal poisons, notably that of the scorpion.

The genus Latrodectus is represented in the United States by at least two species, L. mactans and L. geometricus. Concerning L. mactans there are very circumstantial accounts of serious injury and even death in man[B]. Latrodectus mactans is coal black, marked with red or yellow or both. It has eight eyes, which are dissimilar in color and are distinctly in front of the middle of the thorax, the lateral eyes of each side widely separate. The tarsi of the fourth pair of legs has a number of curved setæ in a single series. It has on the ventral side of its abdomen an hour-glass shaped spot. The full-grown female is about half an inch in length. Its globose abdomen is usually marked with one or more red spots dorsally along the middle line. The male is about half as long but has in addition to the dorsal spots, four pairs of stripes along the sides. Immature females resemble the male in coloring ([fig. 9]).

Regarding the distribution of Latrodectus mactans, Comstock states that: "Although it is essentially a Southern species, it occurs in Indiana, Ohio, Pennsylvania, New Hampshire, and doubtless other of the Northern States." L. geometricus has been reported from California.

Other Venomous Spiders—While conclusive evidence regarding the venomous nature of spiders is meager and relates almost wholly to that of the genus Latrodectus, the group is a large one and we are not justified in dismissing arbitrarily, all accounts of injury from their bites. Several species stand out as especially needing more detailed investigation.