FORMS OF SPICULES. 68 



in other words, the complexity is a part of the continuous 

 development of the organism. 



Only in the case of such organisms as Radiolaina, in which 

 the scleroplasm, i.e. the portion of the body-substance in 

 which the deposition of spicule-substance occurs, is con- 

 tinuous with the rest of the organism, do the above argu- 

 ments fail to apply. Grranting this we may still ask what is 

 the evidence that in Radiolaria the forms of the spicules are 

 inherited ? And the answer from most quarters will be : the 

 same class of evidence which leads us to state that the forms 

 and patterns of the siliceous valves of diatoms, e.g. are 

 inherited. The siliceous valves of diatoms and the siliceous 

 spicules of Radiolaria are both cytoplasmic secretions, and 

 surely what applies to one applies to the other. In replying 

 to this criticism, a distinction must first be drawn between 

 the general region of deposition of the silica and the par- 

 ticular distribution of the silica within that region. The 

 general shapes of the diatom valve and the radiolarian shell 

 (or region containing the spicules in those Radiolaria and 

 higher organisms which possess no shell) indicate the former 

 and are undoubtedly inherited — are determined, like the 

 form of bone, by the disposition of the scleroplasm, i.e. by a 

 mould. But, on the other hand, it is altogether another 

 question as to whether the forms of the individual spicules or 

 the patterns on the radiolarian shell or diatom valve are 

 inherited.-^ Anyone who questions the inheritance of these 

 patterns is at once met with the answer that particular pat- 

 terns are constantly reproduced by particular organisms, and 

 that therefore they must be inherited (!) — an answer alto- 

 gether beside the mark. Given a colloid of the same compo- 

 sition under similar conditions, and if it produces a particular 

 pattern on one occasion for pui'ely physical reasons (onto- 

 genetic causes) it will on another, just as calcite crystals are 

 always produced whenever the required conditions are con- 

 formed to. Previous to the experiments of Herbst, the arms 



* Tlie patterns on dinoflagellate and desmid plates, coats of pollen-grains 

 and seeds, and scales of ganoid fishes may also perhaps be included here, 



