SECRETARY'S REPORT. 199 



oflen meeting each other on their way, g-oing- over very nearly 

 the same ground. 



The returns of the United States census are grossly inaccurate, 

 and untrustworthy from beginning to end, and afford a most 

 unsatisfactory basis for forming a correct judgment of the pre- 

 sent condition of the agriculture of the State ; but the returns 

 of the assessors, though themselves far from perfect, are more 

 accurate and worthy of confidence, and afford a valuable means 

 of correction in many cases. 



The worthlessness of the United States census is probably 

 owing to two prominent causes, which will vitiate this and 

 every subsequent census till these causes are removed. The first, 

 is the carelessness and apparent want of interest in making 

 out the blanks at head-quarters, as if the inquiries were of no 

 consequence. This arises, perhaps, in a great measure, from a 

 well-founded conviction of the utter worthlessness of the re- 

 turns when they are made, owing to the second cause, which is 

 the manner in which the statistics are collected by the assistant 

 United States marshals all over the country, who, one would 

 suppose, are appointed wholly without reference to fitness for 

 the work. 



The immense patronage and the boundless political favor- 

 itism connected with this as well as other branches of the 

 public service will probably ensure its continuance, or at least 

 make it extremely difficult to break it up, and substitute some 

 more efficient system in its place. But it would seem as if 

 the statistics might be returned by the assessors of towns, the 

 county commissioners, or, in cases where these officers do not 

 exist, by the boards of supervisors of counties, or other officers 

 chosen or appointed for their fitness to make these inquiries, 

 in which case they would be worthy of greater confidence. 

 Perfection is, of course, out of the question, but an approxima- 

 tion to accuracy is certainly attainable. A reference to the 

 table on page 299 of the statistics, and the instances cited in 

 the course of the remarks connected with them, together with 

 innumerable other examj^les which might be stated, would 

 lead to the belief that the returns of the eighth census cannot 

 be regarded as even a remote approximation to the truth. 



But imperfect as these statistics are, they are still the best 

 we shall have for the next ten years, and, as such, are extremely 



