4 



the propriety of this is clearly seen iipon a moment's reflection. The relation 

 between production, interchange, and consumption hind all together, for they 

 embrace supply and demand, and it is these which give value to all commodities, 

 and according to their value will be their production. It has been well said 

 that the great interests of agriculture, manufactures, and commerce are one aTtd 

 indivisible ; that they are intertwined and dependent upon each other. All 

 national industry must be regarded as a luiity, and hence Mr. Webster correctly 

 declared that the cotton spinners of Lowell were growers of cotton, because 

 they supplied clothing to those who worked in the cotton fields. And the laborer 

 carrying the cotton cloth of the one to the other, and the cotton of the latter to 

 the former, is, for like reason, a producer also of cotton. A Department having 

 cognizance of the interests of the one, must necessarily have like power over 

 those of the other. 



Whilst it is certain, then, that this department must ultimately have this extent of 

 power and duty, yet as a new Department should extend its operations gradually, 

 all that is noio needed from this indivisibility of agriculture, manufactures and com- 

 merce, is the general power to collect statistics by circulars of all these, as it 

 now exercises over the statistics of agriculture. 



But whilst this extent of power is all that is needed «ow, it does not constitute 

 any part of the basis on which estimates of crops, &c., are made. This is found 

 only in the census returns. The estimate of crops from year to year will 

 always be by comparing them with those of the preceding year, and these are 

 ascertained by the census. The first basis, therefore, is the census. Its extent, 

 like the statistics obtained each month by circulars, should embrace the leading 

 products of all our industrial piirsuits. Hov often should this basis be reniwed, 

 or, in other tcords, how often should the census be tahen ? 



2. The experience of the last summer and fall shows that a census should be 

 taken every five years. This w^ould be often enough. For the other four years 

 estimates can be made through the information furnished by correspondents. 

 But the general government orders but one census in a decade, and the inquiry 

 is, whether any means may be relied on to obviate the necessity of another one"? 



If the several States could be induced to adopt a uniform mode of taking a 

 census every ten years intermediate with that of the general government, the 

 basis given by them would be sufficient. But few of them take any census at 

 all, and those that do have reference more to their own industry than to the as- 

 certainment of statistical facts, essential to the development of general matters, 

 such as the internal trade of the whole country. To the accomplishment of 

 such a purpose all the statistics appertaining to agriculture, manufactures, and 

 commerce, must be taken by a uniform mode as to qualifications of the persons 

 taking it, as to the directions for their government, as to penalties for failure of 

 duty, and those imjiosed on persons withholding the information sought. It 

 would be needless to hope for such united action in so large a number of States 

 as compose the Union, especially when but few have ever as yet had a State 

 census. 



The practical question, therefore, must be this : Will the general government 

 take a census every fifth year ? Certainly not, if the delays and expenses are 

 the same as of the census taken every tenth year. 



Two things are necessary to avoid them. First, a system well devised to the 

 special objects it aims to accomplish; and second, agents who are skilled in their 

 duties, and bring to their discharge a proper appreciation of the object of the 

 work in which they are engaged. 



The utility of a census lies in the principles which its statistics embody. 

 These principles should be well considered before the objects embraced in the 

 census are selected, because they should have strict reference to unfolding these 

 principles. Thus, for instance, if the development of the internal trade, Avith 

 all its bearings on the industry of this country and the political relations to 



