"...the team suggests that frequently quoted iridium values are incorrect. Using a comparison with another extraterrestrial element deposited in the impact - osmium - they were able to deduce that the collision deposited less debris than has previously been supposed...The recalculated iridium value suggests a smaller body hit the Earth."<p>I don't know if this instance is the case, but I have noticed a few instances where false knowledge is repeated for years. The ones that I have noticed have been due to researchers being lazy or too rushed to look at the primary source. It always makes me wonder what else we have not yet corrected. The viscosity incident that Feynman mentioned, while different, also freaks me out. See: <a href="http://en.wikipedia.org/wiki/Oil_drop_experiment#Millikan.27s_experiment_as_an_example_of_psychological_effects_in_scientific_methodology" rel="nofollow">http://en.wikipedia.org/wiki/Oil_drop_experiment#Millikan.27...</a><p>Example:
A cup of green tea is often said to contain 125-250mg of egcg. While this is true for a gram of sencha or matcha, where the whole leaf is ingested, a hot water brewed tea provides more than an order of magnitude less. Sencha and matcha were measured by ethanol extract, simulating leaf ingestion. But for about a decade now, the original source is not cited but other papers claiming water brew to contain >100mg. If you follow the trail all the way back to the original source, you can see that one person misread the paper and started this false knowledge. It is something I think about a lot when validating my own references.