The speech peculiarities of dreams have been very thoroughly investigated by Kraepelin,[34] who has brought together two hundred and eighty-one examples, partly observed in himself, though they are not common, and Kraepelin considers that the hearing centres fall more deeply asleep than the visual centres, the eyes being already sufficiently protected by the lids.[35] Kraepelin classifies the speech disturbances of dreams into two great groups: (1) _paraphasia_, or disturbance of word-finding, where the idea is associated with a wrong word, which is sometimes a new formation[36]; and (2) _disorders of oration_, in which the peculiarity lies, not in the words, but in their order. The speech disturbances of dreams, Kraepelin remarks, spring from deep disturbance of thought, such as occur in sensorial aphasia, and, as in such aphasia, the dreamer thinks his nonsense is quite clear and reasonable. Much the same may occur in alcoholic delirium and in _dementia præcox_. The invention of new words probably occurs frequently in dreams, without leaving a clear trace in memory, and still more frequently, perhaps, as in the 'selvdrolla' dream, already recorded, there are seeming new verbal formations which are really mere corruptions of imperfectly realised words. An example of a definite and precise new word seems to be furnished by the following dream, which was at all points vivid and precise. I saw quite close to me a huge tawny bird, with an orange bill. The creature got up and moved away, seeming as large as an ostrich. I asked a lady, standing by, who had some ornithological knowledge, what the bird was, and she replied that she thought it was a _jaleisa_. Then I asked the same question of a poor woman who was passing, curious to know what she would answer; she said, 'Oh, it's a kind of starling.' There was no doubt in my dream as to the spelling of 'jaleisa,' but I am quite unable to account for the word.[37] It so happened, however, that before I went to bed I had been reading one of Calderon's plays, and I imagine that this pseudo-Spanish word had formed itself in my brain among the echoes of Calderon's enchanting music. The question arises as to why that ignorant old woman should have called the jaleisa a starling. It seems just possible that the more familiar name was suggested by the last syllable of the strange bird's name, the association being verbal. It is equally possible, and perhaps more likely, that the association followed by the more usual visual channel, and that the jaleisa's orange beak suggested the large orange beaks of newly hatched starlings, which had once, many years previously, vividly attracted my attention.