On the naturally occurring glycopeptides of urine and of connective tissue