Much valuable information as to the anatomy of Thysanura has been obtained by Grassi and Oudemans, and is of great interest. Taking four genera, viz. _Campodea_, _Japyx_, _Machilis_, and _Lepisma_, to represent the four families constituting the sub-order, we will briefly enumerate some of the more remarkable of the characters of their internal anatomy. _Campodea_ has a very inferior development of the tracheal system; there are three pairs of spiracles, which are situate on the thoracic region; the tracheae connected with each spiracle remain distinct, not uniting with those coming from another spiracle; there are thus six separate small tracheal systems, three on each side of the body. _Japyx solifugus_ has eleven pairs of spiracles, of which four are thoracic; the tracheae are united into one system on each side by means of lateral tubes; thus there are two extensive tracheal systems situate one on each side of the body, there being a single transverse tube, placed near the posterior extremity, uniting the two lateral systems. In _Machilis_ there are nine pairs of stigmata, two of them thoracic, seven abdominal; the tracheae from each spiracle remain unconnected, so that there are eighteen separate tracheal systems, some of which are considerably more developed than others. The Lepismidae have ten pairs of stigmata, and the tracheae connected with them are completely united into one system by longitudinal and transverse tubes. Besides these differences there are others, of considerable importance, in the position of the stigmata. {187}All the Thysanura possess salivary glands. In _Campodea_ there are about sixteen extremely short Malpighian tubules, or perhaps glands representing these organs; _Japyx_ is destitute of these structures; _Machilis maritima_ has twenty elongate tubules; in _Lepisma_ also they are long, and apparently vary in number from four to eight in different species. The proportions of the three divisions of the alimentary canal differ extremely; there is a very large proventriculus in _Lepisma_, but not in the other families; coecal diverticula are present on the anterior part of the true stomach in _Machilis_ and in _Lepisma_, but are wanting in _Campodea_ and in _Japyx_. The dorsal vessel seems not to present any great differences in the sub-order. Grassi says there are no alary muscles present, but Oudemans describes them as existing in _Machilis_, but as being excessively delicate. The ventral chain of nerve-ganglia consists in _Campodea_ of one cephalic ganglion, one sub-oesophageal (which clearly belongs to the ventral series of ganglia), three thoracic, and seven abdominal. In the other families there are eight instead of seven abdominal ganglia.