" I have quite a bit of confidence in the USDA lab at Gastonia."

That lab reported glucose at 15.58%, fructose at 16.18%, disaccharides at 13.86% and trisaccharides at 14.04%.  I have never seen any honey with such high trisaccharides.  But, lets assume it is correct.  Then the total solids content of this sample was 59.7%.  If that is correct this stuff should pour like water.  Or the other explanation is there is some massive amount of some other sugar or even non sugar in the sample that was not part of the standard sugars analyzed. I am not impressed when they report disaccharides instead of sucrose.  They are running a mass spec detector so could easy enough tell if that component was sucrose or some mix of sucrose with other disaccharides.  If all they are looking at is total ion current than the analysis for disaccharides is close to meaningless as         equi-molar amounts of two disaccharides can have very different total ion currents.

On the other hand the commercial lab reported glucose at 28.69%, fructose at 35.77% and sucrose at 15.09% which add to a total solids content of 79.6%.  This is still low for total solids on a honey sample, but only by 3%, which i suppose could easy enough could be made up by sugars not analyzed.  Or possibly it is even correct.

Personally I am not impressed with either set of numbers.  At best they are both reporting one more significant figure than justified.  I see no reason from the data as reported to believe either one is right.  The ratios of sugars reported by the two sources do not even come close to being equal so you can rule out a mistake due to a dilution error before analysis as causing the huge difference between the two labs on total solids.  If forced to pick one as being correct I would pick the commercial lab unless that honey pours like water.  If it pours like water I would pick the USDA results after I did a refractive index measurement to get some idea that total water was in the 40% range.

Such interlab disagreement in results comes as no surprise to me at all.  My lab was involved in enough round robin analyses to measure how well methods worked for me to have seen discrepancies this big and even a lot worse.  There are lots of good labs out there that do a great job nearly every time.  Every single one of them totally blows it sometimes including the one I ran when I ran it.  And there are a bunch who provide cheap numbers for those who simply need a number to fill in some blank and who really do not care at all if the number is really right or not.  You would be surprised how many just need a number to fill in a blank.  They are the majority of all samples sent to labs.  I recall a VP asking me once what % of the numbers we reported really helped the sample submitter.  My answer was probably in the 1% range and almost all of those were the result of getting a surprising result that causes the submitter to think hard about what this result really meant to his project.

Dick

             ***********************************************
The BEE-L mailing list is powered by L-Soft's renowned
LISTSERV(R) list management software.  For more information, go to:
http://www.lsoft.com/LISTSERV-powered.html