58 



DB. W. EUTHEEFORB ON THE (ESOPHAGUS 



B, a' and b', a" and b", a'" and b'", the bands form the outer layer, which, 

 however, ia structurally complete only between a' and b', a" and b". The bands 

 form the inner layer between b and a', b' and a", b" and a'". The inner 

 layers, unlike the outer, are, on the model, all perfect ; it will be observed that 

 they contain only single bands, and are, in consequence, only half as thick as 

 the outer layers between a' and b', a" and b", which have double bands. 



The whole could be converted into a solid tube by multiplying the bands, 

 thereby filling up the open spaces. 



Just "before the gullet joins tlie stomacli, the inner layer be- 

 comes very much thicker than the outer layer, owing to the 

 presence of circular fibres, which probably act as a sphincter. 



Two other theories suggested themselves to me in the course of 

 my investigations : — First, that although a perfect loop, the bundle 

 might be longer than I now suppose it to be ; that is, instead of 

 crossing the oesophagus only thrice, it might do so ofbener, say 

 five or seven times (Fig. 8). This I soon abandoned; for I 



Fig. 8. 



found that all the fibres of the inner layer, after passing into 

 the outer, become non-decussating at the next decussation,-which 

 could not be the case if the loop crossed the gullet more than 

 three times ; and, moreover, I found that at a decussation the wow- 

 decussating bundle is as large as the decussating bundle, which 

 could not be the case were this long-loop view correct. 



The same facts overthrew the other theory which suggested 

 itself to me, viz. that the fibres form long bundles, beginning by 

 loops at the decussations — at one decussation the convexities of 

 the loops pointing downwards, and the bundles running spirally 

 to the upper extremity of the oesophagus, while at the other 

 decussation the convexities of the loops point upwards, and the 

 bundles run spirally towards the lower extremity. According to 



Fig. 9. 



this view, the gullet would certainly have been of equal thickness 

 throughout ; but its truth is negatived by the same facts which 

 oppose the long-loop view, viz. that the non-decussating bundles 

 at the decussation are as large as the decussating bundles, and 



