PATHOLOGY: W. S. HALSTED 
205 
in man. We found, in dogs, as was to have been expected, that fine, completely 
occluding, ligatures (sizes C or E sewing silk) applied to the thoracic aorta 
just below the arch would cut through in about two days, and invariably 
with promptly fatal hemorrhage; whereas coarse ligatures usually made their 
way through the aorta wall very slowly and without leakage of blood. A 
connective tissue diaphragm often forms in the wake of these broader threads 
and the lumen of the vessel may be more or less completely reestablished. 
It occurred to me after much experimentation that occlusion of the aorta 
to a degree not sufficient fatally to overburden the human heart might effect 
the cure of an aortic aneurysm. Knotted ligatures we found to be unsuitable, 
for a desired degree of constriction or obliteration could not be accurately 
obtained nor could the crushing of the arterial wall be invariably avoided. 
Tapes of various materials were tested — of cotton, of chromicized intestinal 
submucosa, of elastic tissue obtained from the aorta, of aponeurotic white 
fibrous tissue. These were applied in spiral or cuff form. Best suited to the 
purpose were bands of metal, of aluminum, accurately rolled in cylindrical 
form by a little instrument of this kind (exhibit). In the use of these metal 
bands it was impossible to crush the arterial wall, and the desired amount of 
obturation could be obtained with precision, and also maintained. 
The infolded and snugly opposed intimal surfaces under the compressing 
band have in no instance adhered to each other, and for the reason that the 
pressure necessary to produce even a very slight reduction in the lumen of the 
vessel has, in my experience, invariably caused atrophy of its wall. When the 
occlusion is complete the necrotic arterial wall included in the metal band be- 
comes replaced by a solid cylindrical cord of fibrous tissue, the substitution 
taking place from the ends. 
An interesting incidental observation which we have made in the course of 
our experiments with the metal band is this; that a dilation of the artery occurs 
just below a band when the degree of constriction is of the proper amount. 
This observation apparently explains in a measure the occurrence of aneu- 
rysms of the subclavian artery distal to a cervical rib. Analyzing 525 clinical 
cases of cervical rib we found 106 in which the subclavian artery had been 
compressed, and that in 21 of these, aneurysm or dilation of this vessel distal 
to the site of constriction had been noted. 
As to the cause of these aneurysms, five of which have come to the knowl- 
edge of the collators, there has been varied conjecture. Commentators are, 
however, agreed that the occurrence of the dilation would have been less in- 
comprehensible to them had it manifested itself on the proximal instead of the 
distal side of the compression. Attempts have been made to explain the 
phenomenon, and the following suggestions offered as to its possible cause: 
(1) Weakening of the wall of the subclavian artery from erosion by the rib. 
(2) Variable or intermittent pulse pressure occasioned by the normal excur- 
sions of the arm. 
