THE PHYSICIAN who attempts to establish etiology by morphology, either macroscopic or microscopic, finds his position increasingly untenable. In other words, the concept that clinical or histopathologic similarity necessarily establishes a causative or nosologic relationship between dermatoses is tottering, although articles grouping cutaneous entities based on such resemblances are still being published. Obviously, pityriasis rosea may resemble secondary syphilis, but the causative factors are different. The same may be said for psoriasis arthropathica and keratosis blennorrhagica, herpes simplex and impetigo contagiosa, contact dermatitis and atopic dermatitis or generalized neurodermatitis. Actually, one must consider many factors to say that a cutaneous lesion represents a hypersensitivity or an infection or a combination of both. Simple inspection and ordinary biopsy studies often fail to establish the nature of a dermatosis. Admittedly, sarcoid, or if one prefers Rostenberg's* term of sarcoid reaction, has been a particular problem in this