Although a perfect reproducibility was not obtained, this paradoxical phenomenon was generally consistent in different observations. Correlations were found between details of laboratory handling, as well as environment temperature, and the size of the results.

(A) In the author's own experiment, there was a clear trend of T30x animals developing more slowly (i.e. up to 6 h within 3 days) than W30x animals. This is in line with the previous experiments. Due to small numbers of animals, the differences in the frequency of larvae reaching the 4-legged stage and the stage with reduced tail were not statistically significant (p > 0.05). The effect size was large (d > 0.08). (B) In the analysis of all available data with regard to the 4-legged stage, pooled T30x values from the initial team were 10.1% smaller than W30x values (100%) and pooled T30x values from the 5 independent researchers were 12.4% smaller (p < 0.01 and d > 0.08). Analogously, the number of animals entering the juvenile stage with reduced tail was smaller for T30x than for W30x.

Introduction: In the course of more than two decades of experimental work on a model with amphibians and extremely diluted thyroxine, one experiment in particular, investigating the effect of an ultra-high dilution of thyroxine (T30x) v analogously prepared water (W30x) in amphibians from highland biotopes, was found to be reproducible. A total of 22 experimental runs were performed between 1990 and 2011, 15 by the initial researchers and 7 by altogether 5 independent researchers (1-5). In most of these (the sole exception being two performed and reported by the initial team) a trend was found of T30x-animals developing more slowly than W30x-animals. Pooled T30x values obtained by the initial team were 10.1% lower than W30x values (100%) (p < 0.01), and pooled T30x values from the 5 independent researchers were 12.4% lower (p < 0.01). The purpose of this study was to test the hypothesis that storing the animals at 4°C prior to the experiment does not influence (i.e. inhibit) the effect of T30x. Cooling here seemed to be a promising means of facilitating the transport of the highland larvae to laboratories and of synchronizing experiments. Methods: The original protocol was followed, but animals were stored at 4°C for several days prior to the experiment. Results: In contrast to the majority of previous experiments, no clear trend was found of T30x values being different from W30x values, i.e. of animals developing more slowly under the influence of T30x (p > 0.05). Conclusion: This experiment failed to reproduce the previously observed inhibiting effect of ultra-high diluted thyroxine on highland amphibians. The hypothesis that storage of the animals at 4°C does not influence the effect of T30x could not be proven; in contrast, it may be that this intermediate cooling down of the larvae is responsible for the failure of the replication.

No abstract