This is included in Saccardo as part of Xylaria, but we feel is well entitled to generic rank. It was proposed by Ehrenberg in 1820 for a curious species collected in Brazil. The genus differs from Xylaria in having the fruiting bodies on the ends of branches, which in one species are dichotomous, or in the other two species sessile or subsessile and borne on a slender rhachis. There are conflicting accounts of the structure of these bodies. The original, by Ehrenberg, represents them as hollow bodies, with the perithecia imbedded in the walls. That also is as shown by Cooke and is the usual idea. Moeller, on the contrary, represents each body as a perithecium, and our examination confirms Moeller's view. If Moeller's account is true, as it seems to be, it is a strong reason why Thamnomyces should not be classed with Xylaria.

The usual Xylaria has a white, sterile, central portion known as the stroma, bearing a carbonous crust. The perithecia are generally imbedded in the outer portion of the stroma, the mouths opening through the carbonous crust. The walls of the perithecia are carbonous, and confluent with the crust. The genus Thamnomyces has a slender stem, entirely carbonous. This seems to have been the main difference between it and Xylaria in the old classification, but the character is fallacious.

There are Species of Xylaria that have no white stroma. The stem is slender and carbonous and bears the carbonous fruit bodies, superficial, but sessile and globose. Fries proposed for these species, the generic name Rhizomorpha, which Saccardo united with Thamnomyces as a section of Xylaria. In my view it is an entirely different idea from Thamnomyces and should form a section in itself in the genus Xylaria. There are Several species like Xylaria scopiformis that intimately connect Rhizomorpha with Xylaria.

We believe the genus Thamnomyces, in the true sense, embraces only three species as follows:

THAMNOMYCES CHAMISSONIS (Fig. 850).—Stem Carbonous, black, smooth, repeatedly dichotomously branched, the ultimate branches bearing ovate, acute fruiting bodies. Structure of these bodies shown by Moeller is entirely carbonous, hollow, each forming a single, carbonous perithecium. Spores shown by different authors as of different shapes and sizes. In our specimens they are 9×20-28 mic., dark, and arctuate. They closely resemble the ordinary Xylaria spore.

This was originally named from Brazil by Ehrenberg, who gave a good illustration of it. It has therefore escaped all synonyms, excepting by Cooke, who discovered it was a new species and called it Thamnomyces dendroidea. Hennings also discovered it from Africa, first as a new variety, then as a new species, Thamnomyces camerunensis, but of course everything that came to Hennings must be "new" something. It grows on rotten, hard wood, and does not seem frequent in our American tropics. In Africa, however, I judge it is more abundant as numbers of African collections are in the museum at Berlin. We have only received it once, at nice specimen (Fig. 850) from R. H. Bunting, Gold Coast, Africa.

THAMNOMYCES CHORDALIS (Fig. 851).—Stem long, slender, several proceeding from a common base, entirely carbonous, black, smooth. Fruiting bodies (or perithecia?) sessile along the stem, ovate, with slender apices, black. Spores oblong, arctuate, dark.

This, I believe, is only known from tropical America, but is apparently not rare as it is recorded a number of times, mostly from Brazil. Fries named it from French Guiana in 1830 and gave a characteristic description of it. A co-type with the fruit mostly gone is at Kew. Later Montagne got it also from French Guiana and gave a good figure and description under the name Thamnomyces rostratus. He thought it was different from Fries' species on account of the spores not being globose, but the "globose" spores of the original description is doubtless an error. The plants are surely the same. As Montagne's figure is characteristic, the plant when subsequently found has usually been recorded under his name. We present in our figure both Montagne's and Fries' type.

THAMNOMYCES FUCIFORMIS (Fig. 852).—In general appearance, this is the same as Thamnomyces chordalis, but a much larger plant. The fruit bodies (perithecia?) are more slender and are short, stalked. Our figure, which is about half the spike, will show exactly the difference between the two species. The plant was named by Berkeley from specimens collected in Brazil by Spruce, and to this day is only known from this old collection. The name is from the habits, "those of a fucus rather than a fungus," a far-fetched comparison, for my impression is there are no fuci that are carbonous, or have much resemblance to this plant.

RELATED PLANTS.

The following plants are compiled in the section Thamnomyces in Saccardo. None of them are true Thamnomyces, and most of them could go into Fries' genera Rhizomorpha. I do not believe, however, it is possible to keep Rhizomorpha separate from Xylaria. The type species Xylaria setosa is quite different from the normal type of Xylarias in having entirely carbonous, filiform stems and superficial perithecia, but both of these features merge into Xylaria through so many intermediate species that there is no drawing the line of demarcation.

XYLARIA SETOSA (Fig. 853).—Stem densely fasciculate, filiform, black, entirely carbonous. Perithecia ovate, sparse, rarely developed. Spores (teste Fuckel), ovoid, dark, 10 × 16 mic.

This is a rare plant in Europe, growing on old sacks, matting, carpets, and similar refuse. It is generally found in cellars. I think it is not known on wood nor recorded in the United States. It resembles carbonized horse hair and was called "horse hair usnea" by old Dillenius. Our photograph of the specimen at Kew will give a good idea of it, although from the account it grows erect, and is not matted. Both Bulliard and Sowerby gave characteristic figures, both from plants growing in cellars, on old mats. It has had a great number of names, and is recorded in Saccardo as Xylaria hippotrichoides, the specific name proposed by Sowerby and used by Persoon. Some very recent juggler, I have forgotten who, dug up the old name setosa, which I adopt as being less cumbersome. Occasionally these jugglers do propose some improvement in names, and I believe in encouraging them, when their wonderful date dictionary discoveries are really better names. Saccardo gives the following synonyms: Sphaeria hippotrichoides, Ceratonema hippotrichoides, Hypoxylon loculiferum, Rhizomorpha tuberculosa, Cryptothamnium usneaeforme, Rhizomorpha setiformis, Chaenocarpus setosus, Chaenocarpus Simonini. The date expert must have had quite a job.

XYLARIA ANNULIPES, described and figured by Montagne as Thamnomyces annulipes from Brazil, is same as since named Xylaria marasmoides (Fig. 854) by Berkeley. Berkeley does not mention the rings on the stem as shown so plainly in Montagne's enlarged figure, nor can I note them with a lens on my photograph of Berkeley's or Montagne's types. Spegazzini refers marasmoides as a synonym for annulipes, no doubt correctly. Theissen refers it as a synonym for Xylaria aristata, an evident error. Xylaria vermiculus, recently published from Brazil by Sydow, as "Saccardo n. sp. in litt.," is, both from description and photograph, evidently the same as Xylaria annulipes.

XYLARIA MELANURA (Fig. 855), West Indies, described as Chaenocarpus melanurus and compiled in Saccardo in section Thamnomyces, is evidently same as Xylaria gracillima in sense of Berkeley and Montagne, but not I believe as to Fries. We present a photograph made from LÉveillÉ's cotype.