In 1890, T. C. Chamberlin, the distinguished American geologist and educator, wrote a classic paper entitled method of multiple working The paper has been repeatedly reprinted in various forms (most recently in the May 7, 1965, issue of Science). On numerous occasions I have returned to this paper, struck by what I regarded initially as its rather curious central theme. In brief, Chamberlin a�gues that a scientist should take serious measures to avoid undue parental affection for a newly constructed theory (or paradigm, to use Kuhnian phraseology) ; that such affection too often leads the unwary to embrace concepts that are at best partial, and at worst erroneous, views of reality; and that to achieve the desired goal of objectivity, a scientist should erect a set of alternative working hypotheses that will serve to temper his enthusiasm for his favorite intellectual offspring. On first reading, this theme seemed flawed-how could a scientist generalize effectively, and how could he have the courage to put forth new but possibly radical and unpopular concepts unless he felt, and presumably felt deeply, that he really had discovered the Rosetta stone of his science? Would not such affection add to, rather than detract from, the workings of science? It seems to me now, however, that Chamberlin's message is well taken. To make progress in science, the acceptance of a given set of interpretations is no doubt required, but only up to a point ; the threshold of reasonability can be passed rapidly if one begins to press the data into a preconceived set of paradigmatic pigeonholes. It seems likely that this threshold has been surpassed in recent interpretations of the Precambrian fossil record. If this has in fact occurred, it is probably a function of the newness of the field. With few data available, it becomes relatively easy to construct uncomplicated, plausible, and seemingly correct hypotheses. Certainly, it could be a result of the style of evidence and argument necessitated hy historical science, where the data are always incomplete. Much of the evidence can be interpreted only in a qualitative, rather subjective, fashion, and proof' by analogy is commonplace. A careful reading of the l iterature suggests that