that I would find myself treading on much more dangerous ground. Having gone through the experience. I must now admit that syntactic analysis presents much more intractable problems than phonology. Some information about the method followed in this treatise is to be found in the introductions to the two parts. As far as possi­ ble I have used conventional terms. but not without explaining in what sense they are to be understood in this work. I admit that there is plenty of room for disagreement with my conclusions; but they are conclusions which I arrived at after giving much thought to the problems in each individual case. Even in cases where the reader finds he must agree to differ. I hope he will find the argu­ ments stimulating.