14 INTRODUCTION. 



each areola now rises in a shortly cylindrical projection from ti® 

 surface of the sporophore, carrying with it a hyaline investment, 

 which becomes constricted at the base of the cylindrical process. 

 This constriction is contiaued untU an elongated membranous 

 stalk is formed, bearing at its apex a globule contaioing the 

 protoplasmic matter with its nucleus. The contents' of the globule 

 develops in the course of a few hours into the ellipsoid spore ; this 

 is enclosed in a membranous wall, and is easily detached from the 

 stalk. The gelatinous sporophore dries to a membrane of the 

 frailest structure, and disappears with the first shower_ of rain. 

 The process by which the eight swarm-cells derive their nuclei 

 from the single nucleus of the areolar space' of the sporophore 

 has not been followed; but, judging from analogy, we conclude 

 that a succession of divisions took place from the original nucleus. 

 It appears uncertain how far the changes met with ia the sporo- 

 phores of Ceratiomyxa have an exact parallel in what is seen in 

 the development of the sporangia of the Endosporem. Taking the 

 sporophore as representing the sporangium, we have in both cases 

 a structure developing from the plasmodium and consisting of 

 supporting elements and spore-plasm. In all the Endosporem, so 

 far as has been observed, the nuclei divide by karyokinesis shortly 

 before the spores are formed, and this division is accompanied in 

 many instances, as before mentioned, by the lobing of spore-plasm 

 into masses of two spores' capacity round the dividing nucleus. If 

 the stalked bodies formed on the surface of the sporophore corre- 

 spond with the spores of \h.^EndosporecB, we should expect a previous 

 karyokinetic division of nuclei to have taken place ; a process 

 which has hitherto, however, escaped detection ia stained pre- 

 parations. We should then view the division of the spore-contents 

 oiCeratiomyxa into eight swarm-cells, as corresponding vdth a series 

 of multiplications of a swarm-cell of the Erjdosporem with arrested 

 cell-division. But the whole process requires further careful 

 investigation, and, with the facts already in our possession, there 

 are two other hypotheses which may be suggested as possible. 

 The areolae of the sporophore may represent the masses of two 

 spores' capacity present round the dividing nucleus in many of the 

 Endosporem ; but in this case the masses become encysted and 

 stalked, nuclear division is deferred until the cysts are fully 

 formed, and it is not until these have been placed in water that 

 the cyst- wall is thrown off and the contents divided into eight 

 naked spores. A third and widely different view takes what have 

 commonly been regarded as equivalent to spores in Ceratiomyxa 

 as representing stalked sporangia, arising in great numbers and 

 regularity from the surface of the gelatinous body, which corre- 

 sponds to a branched and complex hypothallus. Each sporangium, 

 which at first contains a single nucleus, on being placed in water' 

 throws off its sporangium-wall and divides into eight naked spores. 

 Should either of the two latter views prove to be the true one, 

 the definition of the Mycetozoa would require to be modified, for 

 the rhythmic streaming of its plasmodium and the character of its 



