The stem attains a diameter of between 2 and 3 feet in the largest specimens; it is made up either of comparatively wide and loosely arranged tubes pursuing a slightly irregular vertical course accompanied by a plexus of much narrower tubes, or of tubes varying in diameter but not divisible into two distinct types. Rings of growth occur in some forms but not in others. Radially elongated or isodiametric spaces occur in the stem tissues in which the tubes are less abundant. Reproductive organs unknown, with the possible exception of some very doubtful bodies described as spores. In 1856 Sir William Dawson proposed the generic name _Prototaxites_ for some large silicified trunks discovered in the Lower and Middle Devonian rocks of Canada. A few years later the same writer[369] published a detailed account of the new fossils and arrived at the conclusion that the Devonian stem showed definite points of affinity with the recent genus _Taxus_, and the generic name suggests that he regarded it as the type of Coniferous trees belonging to the sub-family Taxineae. The reasons for this determination were afterwards shown by Carruthers to be erroneous. Dawson thought he recognised pits and spiral thickenings in the walls of the tubular elements, as well as pointed ends in some of the latter. The spiral markings were in reality small hyphal tubes passing obliquely across the face of the wider tubes, and the apparent ends of the supposed tracheids were deceptive appearances due to the fact that the tubes had in some cases been cut through in an oblique direction. In 1870 Carruthers[370] expressed the opinion that Dawson’s _Prototaxites_ was a “colossal fossil seaweed” and not a coniferous plant. The same author[371] in 1872 published a full and able account of the genus, and conclusively proved that _Prototaxites_ could not be accepted as a Phanerogam; he brought forward almost convincing evidence in favour of including the genus among the algae. The name _Prototaxites_ was now changed for that of _Nematophycus_. Carruthers compares the rings of growth in the fossil stems with those in the large Antarctic _Lessonia_ stems, but he regards the histological characters as pointing to the Siphoneae as the most likely group of recent algae in which to include the Palaeozoic genus.