Experimental techniques and results of a large, very thin vessel loaded on a very small nozzle are represented. These results, as well as others, indicate that analytical results can overestimate nondimensional stress resultants up to a factor of about three for these types of vessels. Stress resultants of large, very thin vessels appear to be independent of the nozzle thickness, when the thickness of the vessel is not much larger or smaller than that of the nozzle. For these types of vessels experimental data may be linearized for each stress resultant, and this linearization greatly reduces the number of diagrams.