Every analyst will admit that none of this even now reads like a statement long since recognized as erroneous. All the same, modifications had to be made as our researches advanced from the repressed to the repressing, from the object instincts to the ego. A cardinal point in this advance was the introduction of the concept of narcissism, _i.e._ the idea that libido cathects the ego itself, that its first dwelling-place was in the ego, and that the latter remains to some extent its permanent headquarters. This narcissistic libido turns in the direction of objects, thus becoming object-libido, and can transform itself back into narcissistic libido. The concept of narcissism made it possible to consider the traumatic neuroses, as well as many diseases bordering on the psychoses, and also the latter themselves, from the psycho-analytic angle. It was not necessary to abandon the view that the transference-neuroses are attempts on the part of the ego to guard itself against sexuality, but the concept of the libido was jeopardized. Since the ego-instincts were found to be libidinal as well, it seemed for a time inevitable that libido should become synonymous with instinctual energy in general, as C. G. Jung had previously advocated. Yet there still remained in me a kind of conviction, for which as yet there were no grounds, that the instincts could not all be of the same nature. I made the next step in _Beyond the Pleasure Principle_ (1920), when the repetition-compulsion and the conservative character of instinctual life first struck me. On the basis of speculations concerning the origin of life and of biological parallels, I drew the conclusion that, beside the instinct preserving the organic substance and binding it into ever larger units,[19] there must exist another in antithesis to this, which would seek to dissolve these units and reinstate their antecedent inorganic state. That is to say, a death instinct as well as Eros; the phenomena of life would then be explicable from the interplay of the two and their counteracting effects on each other. It was not easy, however, to demonstrate the working of this hypothetical death instinct. The manifestations of Eros were conspicuous and audible enough; one might assume that the death instinct worked silently within the organism towards its disintegration, but that, of course, was no proof. The idea that part of the instinct became directed towards the outer world and then showed itself as an instinct of aggression and destruction carried us a step further. The instinct would thus itself have been pressed into the service of Eros, in that the organism would be destroying something animate or inanimate outside itself instead of itself. Conversely, any cessation of this flow outwards must have the effect of intensifying the self-destruction which in any case would always be going on within. From this example one could then surmise that the two kinds of