Implants were made into forelimbs of Triturus viridescens using fresh, frozen and boiled kidney and liver of T. viridescens and R. pipiens. Limbs were recovered at intervals up to 70 days post-implantation.Kidney implants from Wisconsin R . pipiens gave twice as many extensive accessory structures as did Vermont frog kidney. Total induction percentages, however, were similar.Quantitative and qualitative parameters for implant-induction of accessory structures were investigated. The decrease in antigenicity and increased rate of cytolysis of frozen implants resulted in increased similarity between frog and newt kidney in rate and pattern of breakdown and in rates of induction. Modification of rate and duration of the release of the stimulating factor from the implant did not result in induction by liver implants.No evidence was found for any increase in innervation prior to or coincident with blastema formation. Implantation and implant cytolysis may cause hypersensitivity of limb tissues to the normal innervation pattern or trophic stimuli from the implant may act with those from the injured limb tissues to produce growth.The general pattern of host reaction to the implanted material was studied and described.It has been found that the most effective implant-induction of supernumerary limb growth in Triturus viridescens achieved thus far has been obtained by the insertion into a Triturus limb of a piece of freshly biopsied kidney from Rana pipiens of Wisconsin stock (Ruben and Stevens, ' 6 3 ) . Comparable rat kidney fragments in these limbs also induced large host formations, which tended to be complex subcutaneous accumulations of cartilage nodules, rather than organized external limbs projecting dorsad from the "mother" limb (Ruben and Stevens, '62). Rat liver, on the other hand, produced no induction, and, except for one isolated series, frog liver has also been found to be ineffective. Implants of Triturus kidney and liver and of Taricha kidney and liver were all essentially noninducing, eliciting the development of only an occasional small subcutaneous cartilaginous nodule. In fact, homograft and heterograft materials produced very little persistent disruption of the internal tissues of the limb. They cytolyzed very slowly, and substantial portions were still present at the end of the 70-day experimental period.On the basis of inferences drawn from information gathered in this manner, an J. MORPH., 217: 213-228.immunobiological model of post-embryonic implant-induced accessory urodele limb morphogenesis was constructed (Ruben, '60). It was suggested that quantitative as well as qualitative parameters were involved and that the "inducing" factor(s) from the foreign implant would have to be released to the limb in a sufficient quantity and over a sufficient time span if they were to stimulate the dissociation of neighboring host tissues and the accumulation and growth of the pre-blastemal mesenchymal cells. If the implant breakdown were too rapid, the stimulus, though strong enough, might be of insuff...
