CHAPTER V.— COMPARATIVE REVIEW. -ASCOMVCETES. 229 



produce a dense hymenium. The gonidia in the highest state of development are 

 abjointed successively in long rows forming a close colourless mass which covers the 

 cushion, each gonidium having the form of a bent cylinder 60 ^ in length and divided 

 by transverse walls into several members (spore-cells), which may be as many as eight 

 in number. Besides these there are some much smaller, but of similar origin, in which 

 the number of members sinks to two. Each segment-cell of these gonidia may 

 develope in a moist atmosphere into a branched hypha, from single branches of which 

 fresh smaller gonidia are abjointed. If the mycelium vegetating in the tissue of the rind 

 of the tree comes to lie exposed in a moist atmosphere, it sends out numerous branches 

 into the air, and from these also countless small gonidia are abjointed. The size of all 

 these small gonidia sinks by regular gradations to 1.5 /x ; they are all cylindrical and 

 rod-like ; those of medium size are divided by a transverse wall into two segments 

 which often separate from one another ; the smallest are undivided. All down to 

 the size of 2 /* can put out germ-tubes or may also multiply by sprouting. The very 

 smallest have never been observed to put out germ-tubes, but they appear to multiply 

 by division and fission and by sprouting. Lastly, gonidia of the smallest kind are also 

 formed in great numbers by slender branches of the mycelium in the interior of the 

 tissue attacked by the Fungus. All germ-tubes, whether from gonidia or from asco- 

 spores, can develope new fertile mycelia in the proper substratum, that is, in the living 

 rind of a tree ; it is doubtful whether the smallest gonidia can produce mycelia 

 (see section LXXIV). There is also some doubt as to the true nature of certain acro- 

 genously abjointed spores which occur on the sporocarp-bearing portions of the Fungus 

 which we are considering; they appear to belong to parasites of Nectria, and it will be 

 sufficient therefore in this place to refer the reader to Hartig and Tulasne. 



The development of Cordyceps with its great variety of forms will be described 

 subsequently in Division III. 



The cycle of development of the one or perhaps two species included under the 

 name of Pleospora herbarum is particularly rich in forms. These are found in dead 

 and rotting herbaceous plants. The results which will be given here were obtained 

 from plants cultivated in nutrient solutions on the microscopic slide. The mycelium 

 forms (1) the perithecia mentioned above with pericellular compound ascospores : 

 (2) gonidia of three kinds produced acrogenously on filiform gonidiophores, namely, — 

 (a) bicellular or pericellular spores resembling the ascospores, in general shape shortly 

 cylindrical to roundish, with dark-brown thick membranes rough with fine points on the 

 outside, named by Berkeley as a form-species in 1838 Macrosporium Sarcinula and 

 therefore called by Gibelli and Griffini the Sarcinula-form ; they are usually produced 

 singly at the end of the gonidiophore ; (b) the Alternaria-form, classed with the old 

 form-genera Alternaria, Sporidesmium, Mystrosporium, and Polydesmus ; these are coni- 

 cally pear-shaped pluricellular compound spores having a smooth light-brown membrane, 

 and arising at the extremities of the hyphae in long and often branched rows (Fig. 34) ; 

 (c) a form said by Bauke to be a microgonidial form, but of which he gives no further 

 description ; it is not the one known as Cladosporium herbarum and placed by 

 Tulasne with Pleospora herbarum, for this, according to all later investigations, does 

 not belong to this place at all and its genetic connection is uncertain : (3) pyc7iidia 

 (see below, section LXXI, Figs. 118, 119) which appear as intercalary formations on 

 branches of the mycelium. A piece of the hypha consisting of one or several cells 

 swells in the same way as when a perithecium begins to be formed, and its cells 

 divide at the same time meristematically and irregularly by walls inclined in every 

 direction. By this process of growth a small-celled parenchymatous body is formed 

 of many layers, which is round or irregularly elongated in shape and seldom more 

 than 0.2 mm. in size, often much less. This body is at first uniformly dense, but 

 towards the end of its growth a central cavity is formed in it surrounded by the many- 

 layered wall ; this cavity is produced by the cells of the central part ceasing to follow 

 the growth of the outer parts in the direction of the surface and therefore separating 



