560 ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 2 8 



a police dog. A pebble was held in the hand a short time, then 

 thrown among many other similar pebbles 10 to 20 yards away. The 

 dog saw and heard in what direction the stone was thrown, but he 

 could not distinguish it by sight from the other stones. After having 

 smelled the person who threw the stone, Albert searched for and 

 found it many times without making a single mistake. To make 

 the test more complex, six persons threw as many stones at the same 

 time. After smelling one of the six persons the dog invariably 

 returned with the stone thrown by him, showing that the dog can 

 easily recognize individual odors of people. The last test was to 

 determine whether Albert could associate a pocketbook with its 

 owner. Four pocketbooks belonging to four men were laid in a 

 row. Albert, of course, could not distinguish by sight the owner of 

 each purse, but he made no mistake in doing so by smell. He was 

 allowed to smell one of the men and then smelled all the pocketbooks, 

 announcing his decision by barking. These results fully convinced 

 Buytendijk that the employment of dogs in police departments is 

 highly justified. However, as a result of more recent tests by Ger- 

 man police officials, the use of dogs in helping to detect criminals is 

 forbidden in Germany. Critics contend that the clogs were not 

 properly tested. 



In 1926 Lohner, an Austrian physiologist, used pieces of wood, 

 alike in appearance and smell, in testing the olfactory sense of dogs. 

 It was soon learned that a police dog can easily distinguish the Avood 

 handled by a person. While the dog was not looking a piece of wood 

 was held in the hand, then put among other pieces which were han- 

 dled only by means of forceps. The dog was next allowed to smell 

 the hand, and was told to find and to bring the Avood that had been 

 held. Other tests, in Avhich the pieces of wood were deodorized, 

 showed that the dog could distinguish the pieces even if they had been 

 touched with the finger tips for no longer than a second or two. He 

 succeeded even when persons other than the one whose odor was to 

 be identified had handled the wood, and even after artificial odors 

 had been daubed on the wood, although he barked his disapproval. 

 This police dog was able to identify the odor of a person when the 

 wood had been put on diiferent parts of the person's body. 



To summarize, dogs, like people, can detect odors of various kinds 

 and qualities and, in addition, are able to analyze a combination of 

 smells and attend exclusively to one component of it. This is a 

 power that we lack almost entirely. When Ave experience the stimuli 

 of two odors at the same time Ave rarely detect both of the qualities 

 in the mixture; usually one of them overcomes or masks the other, 

 or a ncAv odor unlike either results. The dog, and perhaps many 

 other animals, can analyze a fusion of smells as a trained musician 



