10 Papers from the Department of Marine Biology. 
can well understand why Coutiére and Hafele were unsuccessful in 
tracing the course of the system. 
Coutiére was nevertheless able to make out some details of their 
distribution. His description may be given in full: 
‘“‘Celles-ci s’ont bien différentes de ce que l’on remarque chez Sacculina 
carcini par exemple; elles sont méme plus réduites que dans le genre Sylon, ou 
Hoeck les a décrites. Ces racines occupent uniquement le bourrelet trans- 
versal du pléosternites dont elles ont vraisemblablement provoqué la forma- 
tion anomale; on n’en trouve nulle trace autour de l’intestin ni entre les 
muscles. Elles envahissent, par contre, les lames concentriques du neuriléme 
externe, tres épaissi chez A. edwardsi et A. avarus. Elles ont un diametre de 
25 u a 30 u et se montrent ramifiées et coutournées en tous sens.”’ 
Though Coutiére is mistaken in supposing that the roots are local- 
ised in the neighbourhood of the sternum, his observations as to their 
absence round the intestine and concentration round the nerve cord 
agree with mine. 
In Pilumnus, Hifele had a particularly unfavourable host for study- 
ing Thompsonia. The external sacs are all situated on appendages 
with an enormously thick cuticle. The only possible method of study- 
ing the root system was by prolonged treatment of the appendage in 
Perenyi’s fluid to soften the chitin, after which sections could be cut. 
From an examination of these he concluded that there was a root 
system of an exceedingly simple kind. Kriiger from a more careful 
examination of material from the same source, overstaining his sections 
and washing out carefully, was able to demonstrate a root system of a 
normal kind and show that the chromatin-rich nuclei which Hafele 
had taken to belong to the root system were actually in the connective 
tissue and blood cells of the host. The root system, then, Kriiger 
decides, does not support the claim that Thompsonia is primitive, and, 
on the other hand, the omission of the Nauplius stage is a mark of 
specialisation. This is perfectly correct, and if he had suspected that 
the roots of adjacent sacs were continuous Kriiger would have been 
able to complete his chain of reasoning. 
My own conclusion is that there is a single root system continuous 
throughout the host, from which all external sacs are budded off so that 
each host is parasitised by a single individual and not by a hundred or 
more gregariously inclined Rhizocephalans. 
I was first able to see the root system in the abdominal appendages of 
an infected Synalpheus. Those which bore external sacs were cut off 
from the living animal and examined under a low power of the micro- 
scope. The endopodites and exopodites are greatly flattened and the 
cuticle is thin and unpigmented. The whole organ is thus transparent 
and the roots are visible as slender strands to which the presence of 
innumerable highly refringent yolk granules gives a greyish colour. 
Usually a single root strand entered each ramus and this gave off a 
