XCVJII BOWMAN LECTUEE. 



manifest only in the solitary one who was reached by 

 some efficient determining cause. In others deafness or 

 mental defect, easily omitted from the record unless 

 directly sought for, may take the place of the eye 

 disease (as in Fig. 39), and thus partly or wholly restore 

 the continuity. 



In respect to consanguinity we have to confess that 

 but few of the records tell us the source and kind of 

 cousinship, an omission that may very much lower their 

 value. In future when taking the family histories of 

 persons whose parents were cousins, it will usually be easy 

 to record whether they both belonged to the affected side 

 of the genealogy or not, and whether they were children of 

 two sisters, or of two brothers, or of a brother and sister, 

 or of a sister and brother. 



Of the persons seen at all ages with retinitis pigmen- 

 tosa a considerable majority are males (at least sixty males 

 to rather less than forty females). This fact may, as 

 suggested before, be an expression of some wider law ; 

 but two other interpretations suggest themselves for the 

 time being, viz., either that the females in these families 

 die in excess before they are old enough to show the 

 disease, or that the malady occurs most in the families 

 that contain an excess of male births. It may be of 

 interest to note that a marked excess of males is also seen 

 in the chronic renal diseases, and in diabetes, whilst the 

 reverse is found in the interstitial keratitis of congenital 

 syphilis. 



Next to heredity and consanguinity comes the influence 

 of ill-health in bringing out a liability to retinitis pig- 

 mentosa where, but for such an exciting cause, it might 

 have remained latent. Probably such an influence may 

 sometimes explain the solitary cases. Of such determin- 

 ing causes some of the acute exanthemata seem to be the 

 commonest, but probably tubercle and syphilis and in 

 rare cases even severe loss of blood may have the same 

 effect (Fig. 38). We may suppose that anything capable 

 of damaging the arterioles might determine the onset of 



