243 



TABI^E No. ,S5 

 DATA OF TABLE 34 CALCULATED TO PERCENTAGES OF ORIGINAL WOOD 



Determination 



Wood 



Parts per 



100 



Hemicelluloses. 



Cellulose 



Lignin 



Furfural yield. . 



6.02 



54.74 



30.60 



5.37 



Series I 



Series 1 1 



Borings (1 ) 



Parts per 



56.1 



3 62 



11.79 



30.60 



3.32 



Borings (2 ) 



Parts per 



,S4.7 



5.10 



10.99 



30.60 



3.22 



Wood 



Parts per 



100 



14.23 



47.45 



27,84 



5,90 



Borings 



Parts per 



50.9 



6,20 



10 96 



27,84 

 4,26 



Reference has previously been made to the possibility that the change in the 

 composition of the wood may occur after the material is ejected by the borer. The 

 compositions ol the first and secoiul samples of borings in Series I furnish some ex- 

 perimental evidence on this point. Although the second sample was purposely left 

 in the sea water for at least six months longer than the first, there was no significant 

 reduction in carbohydrate constituents (see Table M). In \iew of the extreme resis- 

 tance of wood to purely chemical action, and the absence of known cellulose-destroying 

 fungi or Ijacteria in sea water, it seems more reasonable to ascribe the loss of cellulose 

 and other carbohydrates to digestive action while in the body than to externa! 

 agencies. 



Further exideiice of the digestion of cellulose was found in the serial sections 

 that were prepared in the course of the investigations of the morphology of Teredo. 

 The contents of the stomach, caecum and intestine, usually remained in place in the 

 sections, as shown in the photomicrographs in fig. 79. When sections were stained with 

 Delafield's haemotoxylin, which is a selective stain for cellulose, the wood fibers in 

 the stomach and in the caecum took the stain, while those in the intestine did not 

 to any appreciable degree. Thus it is evident that at least a part of the cellulose had 

 been remo^•ed from the wood fibers during their stay in the stomach and caecum. 

 Unfortunately, the difference in color does not show in the photomicrographs. Recent 

 work by Miller and Boynton (1926) has shown that Bankia setacea is also capable of 

 utilizing wood as food, wood removed from the caecum having been found to contain 

 about four times as much of reducing sugars as the original undigested wood. 



Considerable evidence has been advanced by Potts (1923) to show that the 

 digestion of wood by Teredo takes place intracellularly, in certain lobules of the liver. 

 As has been stated above, the livers of Teredo are histologically differentiated into two 

 distinct types of lobules, one type (dorsal) being made up of relatively large, close-set, 

 columnar cells, while the ventral lobules consist of smaller, somewhat flattened cells 

 surrounding large lumina. It was observed by Sigerfoos (1908) that these lumina 

 are characteristically filled with particles of wood. According to the observation of 

 Potts, the wood particles taken into the lumina are there seized upon and ingested 

 by free-moving amoeboid cells, within which the wood is presumably digested and 

 prepared for absorption by the animal. Some of the epithelial cells of the liver are 

 also said to contain wood fragments, and it is probable that the free amoeboid cells 

 in the lumen are derived from this epithelium. 



This matter has been further studied by Yonge (1926b), who concludes that 

 digestion is probably largely intracellular in most, if not all, the lamellibranchs (Teredo 

 ■norvegica was one of the forms specially investigated). The so-called livers he finds 

 to be not secretory in function, but organs of absorption and intracellular digestion. 

 Amoebocytes, which are found in the lumen of the gut and in and beneath the epi- 



