OUR interest in liver changes in exophthalmic goitre was aroused last year at the post-mortem examination during the same week of two cases of Graves' disease. Both showed marked atrophy and fibrosis of the liver without signs of passive venous congestion or cardiac failure. This paper attempts to answer the question which arose in our minds, viz. is there a true association between exophthalmic goitre and liver damage independent of venous staais 1Our material consists in all of 30 proved case3 of exophthalmic goitre coming t o mtopsy at U.C.H. during the years 1924-35. Nine of these were personally investigated, the remainder we havo taken from the records of the pathological department of the Medical School. There has been no fielection of cases, except to exclude one with severe gall-bladder infection. Tho series thus includes t i group with obvious possive venous congestion.No case was complicated by chronic infection such as tuberculosis or syphilis. In addition we have collected from the literature records of more than 100 cases with autopsies and the details of those we shall use in our final discussion. Representativo pieces of liver have been blocked in paraffin and sections stainod with Ehrlich's mid hzmatoxylin and watcry eosin and with Weigert's iron hzmatoxylin and Van Gicson. In our own caam and a certain number of the others frozen sections stained for fat and pigment were available.
Materid.Throughout this paper we use certain terms to which we attach very definite meanings.By fat29 change we mean the occurrence in nucleated liver cells of small or large globules of fat, specifying the degree or intensity of the change by such descriptive words aa slight, moderate, marked or extensive. There is no implication of cell damage ; when this is present we state it in unequivocal fashion.By general atrophy we mean that state of the liver characterised by diminution in size and weight beyond what experience has taught us t o expect at a given age period, together with microscopical evidence of smallness of liver lobules and of liver cells. Lobule size we estimate through the closeness of the portal canals a~ compared with that in known healthy livers. We take good care to base our conclusions on representative pieces of liver, avoiding 267 268