Based on the methods from the BS, there is a different procedure used to wash and prepare the sample for a sieve test (i.e. the part greater than 0.075mm) and the sample used in the hydrometer test. With a normal grading the sample is first dried to constant mass, then washed to remove the fine particles. The act of drying can cause the clay fines to bond together and/or change shape and when they are washed they do not readily break down, even when using sodium hexametaphosphate which is supposed to floculate the clay particles. This can result in that some of the material retained on the sieves would pass through the finest if it had not been dried before washing. In conjunction with this, for the hydrometer the sample is prepared from natural, and an identical seperate portion is dried and used to calcualte what the dry mass would have been. This means that in the majority of cases, all the clay particles are able to easily floculate, giving a finer grading.
The result of this is that there can be a discrepency between the percentage passing for the grading using the sieves for the upper part, and the percentage found for the hydrometer.
To get around this, there are a couple of options.
1. A quick, yet not entirley accurate method is to use the percentage passing the 75µm from the grading, in the calcualtion of the particle size for the hydrometer.
2. A better option is to retest the grading, and this time wash the sample from natural with an identical portion used to calculate what the dry mass would have been, had the sample undergone the drying before washing. This should mean that any cementing of the particles will have been prevented.
Both of the above assume that the sample used for the grading and the hydrometer are representitive of the entire sample. It is also not unknown for the sample used in the hydrometer to be 'selected' for the test, which can often produce a sample which is not representitive of all the size fractions within the sample. In this instance the option 1 above would give a better assessment, as although the Hydro will give an accuarte breakdown of the particles passing the 75µm, it needs to be 'matched up' with the overall grading.
Anther thing to consider, have you checked the particle density/sepcific gravity of the soil? Does the calcualtion of the dry mass used in the test make sense?
As stated in your thread, the percentage drops after the 75µm, and to me this would indicate that the sample used in the hydro, was either not fully washed (i.e. once they had sufficient fines for the test, they stopped, which can result in only the finest material passing, with the coarser silts still in the original test portion) or the portion used for the hydro was not representitive.
