CERIANTHARIA. 



absent, wliilst they are characteristic of the median streak in the species bearing acontia, thus tends 

 to shew that the acontia in A. oligopodus and sarsi and in the typical species of the larval genus 

 Arachnactis are differentiations of the median streak, that is, of its continuation, the craspedion 1 ) whilst 

 there is nothing essential to their structure that suggests the cnido-glandular streak. 



If then the acontia in question are differentiations at the aboral end of the craspedion they 

 ought to be homologous with the craspedonemes of the craspedia in Ceriantheopsis amcricanus, 

 though the development of these two organs would have taken a different course. In that case the 

 "acontia" have retained the mucus cells of the median streak and augmented them, and the muscle 

 layer of the streak has not been lost, but on the other hand the descending and ascending limbs of 

 the acontium filament have drawn close together through the reduction of the endoderm part, the 

 acontitim becoming rounded and presenting an oval form in transverse section. The said craspedo- 

 neme from the craspedia, on the contrary, has reduced the number of mucus cells and spirocysts 

 and lost the muscle layer almost entirely, but on the other hand granular gland cells and thick-walled 

 nematocysts have arisen in greater numbers - - just as thick-walled nematocysts have been developed 

 in the most aboral portion of the craspedion in several of the species not provided with acontia (as 

 in C. lloydii) - - though the form of the craspedoneme has undergone no change but corresponds 

 with that of a craspedoneme from the ciliated tract region. 



My study of the "acontia" in Ceriantharia has thus led me to a different conception of their 

 structure and origin from van Beneden's. It is now clear, I think, that van Beueden's views 

 on this point must be considerably modified. Above all we must insist that the principal parts of 

 the "acontia" are ectoderm al (granted that the filaments themselves are ectodermal formations 2 ), 

 which seems highly probable), further that the "acontia" are not sting batteries in the same sense as 

 the "acontia" in Actiniaria, for either the large thick-walled nematocysts are very sparse or if they 

 are found in larger numbers it is in the endoderm part of the "acontium" only. Mucus cells on 

 the contrary are to be found in great numbers. Under these circumstances the propriety may well 

 be questioned of homologizing, as van Beneden does, the "acontia" of Ceriantharia and the acontia 

 of Actiniaria, the latter of which probably have no nematocysts in the entoderm. In my view the "acon- 

 tia" of Ceriantharia have nothing in common with those of Actiniaria, and it would be more correct 

 to give the "acontia" of Ceriantharia another name, such as muco-craspedonemes. At the most 

 we can only speak of acontium-like threads in Ceriantharia, and that mainly with reference to their 

 outward form. I ought to mention besides, before leaving the subject of the acontia in Ceriantharia 

 and Actiniaria, that van Beneden's statement, that their attachment with the mesenteries corre- 

 sponds in the two groups, does not hold good. Van Beneden says that he has many times noticed 

 that the acontia run laterally in Ceriantharia at the point of junction. I have thought I noticed 

 myself an attachment of this kind, but this circumstance depends in my opinion on an irregular con- 



') As the median streak and the craspedion correspond in structure, and the first is continuous with the ectoderm 

 of the stomatodaeum, they are doubtless to be regarded as primary filaments from which the cnido-glandular tract has been 

 developed. 



2 ) Independently of the morphological support given above for the ectodermal origin of the filament we may adduce 

 the fact that the mesenteries in the development zone do not acquire filaments until they have reached the lower border of 

 the stomatodaeum. 



