12 



Papers from the Department of Marine Biology. 



The root system is not localised, then, in the neighbourhood of the 

 external sacs; on the contrary, the main development is in the ventral 

 body wall of the abdomen and often, too, the thorax, where it exists as 

 a network of thin roots underneath and on both sides of the nerve cord, 

 giving off frequent branches into the longitudinal musculature (see 

 text-figure 3). In the thorax the roots extend into the lateral and 

 dorsal body wall as well. Their distribution is not so wide as that of 

 Sacculina, in that the intestine and stomach are not surrounded, but 

 roots occur between the lobes of the liver. In the concentration of 

 roots round the nerve cord rather than the alimentary system Thomp- 

 sonia resembles Mycetomorpha (8), the root system of which has, how- 

 ever, a much more restricted distribution. 



Branching is dichotomous and anastomosis of neighbouring divis- 

 ions certainly does sometimes occur, a feature unknown in Sacculina. 



FIG. 3. Root system of Thompsonia 

 around the ventral nerve cord of 

 Synalpheus brucei. It has been 

 partially torn in removing from the 

 body, and so the network is incom- 

 plete in places. Fixed in Flem- 

 ming's fluid. The root system r., 

 here and in the figures which follow, 

 is revealed by the blackened yolk 

 globules which it contains, v.n.c., 

 ventral nerve cord; l.n., lateral 

 nerve. X38. 



This central or internal part of the root system is concerned, without 

 doubt, with the nutritive function and is loaded with large and numer- 

 ous yolk granules. From the trunk there pass out into the walking 

 legs, abdominal swimmerets and tail-fans, lateral roots which, together 

 with their branches, may be distinguished as the peripheral division 

 of the system. It gives rise to the external sacs and its function is 

 reproductive. The yolk granules, though numerous, are small. 



In the case of the abdominal appendages a single root usually enters 

 the protopodite, but it immediately divides and as a rule both the endo- 

 podite and the exopodite are invaded by secondary roots, which divide 

 further, as mentioned above. Where the roots enter the abdominal 

 appendages there is sometimes also an invasion of the space within the 

 adjacent pleuron. Such a case is figured here (text-figure 4). A com- 

 plicated knot of roots of the nutritive kind occupies the wider part of 

 this space near the attachment of the muscles of the appendage and 

 from this are given off two branches which pursue a tortuous course 



