In view of the clear language of the Court in 82 Fed., 839, and 157 Fed., 849, it is evident that any valuation which does not take into account the appreciation or depreciation of land values cannot be sustained. There can be no serious objection to the doctrine that the property of a corporation generally increases or decreases in value in the same proportion as adjacent property, and it must therefore be admitted that a value based on the sale value of adjacent lands is a reasonable one and must stand. This reasoning, of course, will be subject to exceptions, in the case of terminal properties, docks and water-front properties, and right of way in large cities, but it is believed to be sound when applied to right of way in the country and in small towns and cities. The next question to be determined is whether the increment of value due to the use of the land is a proper one. It would appear that in the use of land for water-works, gas-works, street-car barns, or other isolated tracts of land used for corporation purposes, this increment would be much less than in the case of a steam or interurban railroad, the holdings of which form a continuous and unbroken strip; and, in the case of street railroads, water-works, and like properties, it would be indeed difficult to compute and afterward sustain any considerable increment. In the case of railway properties, however, it is quite evident that the following facts can be sustained: The actual cost of property purchased for railway purposes will range from two and one-half to five times the selling price of similar and adjacent property used for other purposes. While the actual percentage will vary somewhat, as between land in cities and in the country, and as between fully settled districts well served by roads and sparsely populated regions, yet the difference is very marked, and is capable of determination by an examination of the public records. This difference can be determined either by a comparison of railway purchases and other transfers, as was done in the later studies in Michigan and in Wisconsin, or by extending the investigation to include assessed valuations, and using the averages, as was done in the work of Mr. Morgan in Minnesota. In establishing figures for use in a valuation, it would appear to be better to base them on an analysis of actual transfers than to undertake to fix values by any methods of examination and personal appraisal. Enough instances of the wide divergence of expert opinion have been cited to show conclusively that such a method, applied to the thousands of acres of a large corporation, may lead to serious error.