THOMAS m-AL. : AKTERIOVENOUS SHUNT 943On the left side only one lateral bud is present. Within the lower lobe itself the posterior basal bronchus is the final representative of the stem bronchus while the other basal bronchi come off earlier as offshoots.A likely possibility is a foregut diverticulum capturing the terminal bud of the developing bronchus resulting in a communication (Das, Dodge, and Pawcett, 1959). Thus an indcpendent evagination from the foregut presumably attaches itself to and makes a communication with an otherwise normal bronchus. 1%. 3.--Cuse 2. Specimen showing the fistulous tract between the lower-lobe bronchus and the oesophagus.The birth of the idea of a congenital factor in our cases accrued mainly from the classic features seen at thoracotomy. As illustrated in Fig. 2 there were no glands to incriminate them as traction agents. In none of these cases were dense inflammatory adhesions seen which would be pathognomonic of a traction diverticulum of the oesophagus causing an oesophagobronchial fistula. The tract was a well-defined, supple tubular structure, and though there were some adhesions over the bronchiectatic lobe they were conspicuously absent over the diverticulum and the peri-oesophageal region.The late appearance of the symptoms may be explained by the presence of a fold of membrane occluding the fistula in the newborn and breaking down in later life (Jackson and Coats, 1929). Mullard (1954) suggests that the oesophageal orifice could be overridden by a fold of oesophageal mucosa when the bolus passes in front of it, giving some protection against the entry of food particles into the fistula. The changes in the bronchi are obviously the result of pneumonitis complicating the fistulous communication.It is more likely that the late onset of symptoms is due to the interference of the normal mobility of the oesophagus brought about by the progressive adherence of the affected lung to chest wall and mediastinal pleura. When the lung becomes fixed this free descent of the oesophagus will be restricted and the fistula is pulled open; while at an earlier stage when there was free movement of the lung, bronchus and fistula as well as oesophagus descended together, the mouth of the fistula being occluded by a fold of oesophageal mucosa (Mullard, 1954). SUMMARYSeven cases of congenital oesophagobronchial fistula with bronchiectasis are reported. The physical signs and investigations are discussed and the embryology is briefly reviewed. Attention is drawn to the relatively late onset of symptoms. The patient may present only with recurrent or chronic attacks of pneumonitis and unless a careful history is taken and diligent search made the presence of a fistula is likely to be missed. All the cases were treated surgically.