Fourthly, there is a tendency to form a pygidial or telsonic shield, "a fusion of the posterior somites of the body, which is precisely identical in character with the metasomatic carapace of _Limulus_." No crustacean shows metasomatic fusion of segments. Fifthly, a large post-anal spine is developed "in some trilobites" (he refers to a figure of _Dalmanites_). Sixthly, there are frequently lateral spines on the pleura as in _Limulus_. No crustacean has lateral pleural spines. These points may be taken up in order. 1. If trilobites have one appendage-bearing segment in front of the mouth, they are Arachnida; if two, Crustacea. This is based on the idea that in the course of evolution of the Arthropoda, the mouth has shifted backward from a terminal position, and that as a pair of appendages is passed, they lose their function as mouth-parts and eventually become simple tactile organs. Thus arise the cheliceræ of most arachnids, and the two pairs of tactile antennæ of most Crustacea. This theory is excellent, and the rule holds well for modern forms, but as shown by the varying length of the hypostoma in different trilobites, the position of the mouth had not become fixed in that group. In some trilobites, like _Triarthrus_, the gnathobases of the second pair of appendages still function, but in all, so far as known, the mouth was back of the points of attachment of at least two pairs of appendages, and in some at least, back of the points of attachment of four pairs. As pointed out in the case of _Calymene_ and _Ceraurus_, the trilobites show a tendency toward the degeneration of the first and second pairs of biramous appendages, particularly of the gnathobases. They are in just that stage of the backward movement of the mouth when the function of the antennæ as mandibles has not yet been lost. If the presence of functional gnathobases back of the mouth, rather than the points of attachment in front of the mouth, is to be the guide, then Triarthrus might be classed as an arachnid and _Calymene_ and _Isotelus_ as crustaceans. In other words, the rule breaks down in this primitive group. 2. Superficially, the eyes of some trilobites do look like those of _Limulus_, but how close the similarity really was it is impossible to say. The schizochroal eyes were certainly very different, and Watase and Exner both found the structure of the eye of the trilobite unlike that of _Limulus_.