The report oh damages was remitted for the purpose of enabling the parties to show the market value, in a sound condition, of fruit of the same marks and brands as the libelant’s fruit damaged by frost; or, second, in the absence of such proof, the average value of Palermo fruit shipped at about the same time. Upon a furl her hearing before (he commissioner, out of 56 different lots or brands, forming the libelant’s goods, nine lots having the same marks and brands are proved to have been sold in this market in a sound condition at about the same time. To this extent the object of recommitting the matter has boon obtained. Of the nine lots thus capable of direct comparison, eight lots are oranges, which show an average difference between the sound and the damaged fruit of a little less than ton cents a box. Only one lot of lemons was found represented in the sale of sound fruit by the same marks and brands as the libelant’s; which, hv comparison, show a loss of 75 cents a box. There were, however, 26 different lots of lemons in the libelant’s goods, and 30 different lots of oranges; so that of all the 56 different lots there is only one-sixth of the number of which the difference in value in a sound and damaged condition is evidenced by anything approaching direct proof.

As respects the remaining 47 lots of the libelant’s goods, no proof has been introduced of their value in a sound condition, either here or at Palermo; nor any proof of the value in either place of other goods of the same brand; nor has the average value of “Palermo fruit shipped at about the same time ” been shown, except by the testimony of Mr. Second), which shows the prices brought for ail oranges and lemons upon seven steamers arriving during the week prior to and the week after the Boskenna Bay; and that the difference in the average price received for all the sound oranges and the average of Wolfe’s oranges was 13 cents a box, and on the lemons 81 cents a box.

The finding of the commissioner, as T understand it, is based partly upon the general estimate^ of witnesses of what injury frost would do a cargo, partly upon a comparison of 37 other lots of oranges and lemons upon the Boskenna Bay, that had marks and brands wholly different from any of the libelant’s goods, with fruit of similar brands and marks sold between 'March 1st and May 1st, and partly upon the fact that this Boskenna Bay cargo was called a “banner cargo.” I think the result arrived at in these modes is too conjectural and uncertain to he approved. The Colon, 10 Ben. 366; The Rossend Castle, 30 Fed. Rep. 462. The burden of proof is upon the libelant to show his damage, and by the best means that are fairly within his power. The case does not show, as it seems to me, the use of those means that must he deemed available to ascertain most fairly the damage to these particular goods. The sales for a month or six weeks after the sale of the Boskenna Bay cargo were so late in the season that the general range of prices was evidently higher, and that period of comparison should be disregarded. The lots *614selected by each side for comparison and shown in their schedules are evidently picked lots, since the schedule presented by each party embraces none of those shown by the other. A comparison of the libelant’s fruit with other fruit of quite different marks and brands on board the Boskenna Bay, and with other sales of fruit similar to the latter class, assumes that damage to the whole fruit on the Boskenna Bay was the same as the damage to the libelant’s fruit. The proof is not sufficient to warrant this assumption. There,is only very general and indefinite ■evidence that all the cargo of the Boskenna Bay was injured by frost, while there are other strong indications that there was nothing like a uniform injury to all. Some was probably injured much, other portions a little, and some none at all.

Again, there were some 500 different lots of fruit on the Boskenna Bay, and only about 37 of these (other than the libelant’s marks) are selected for a comparison with the sales of sound fruit of similar marks. This is only about one:fourteenth of the whole cargo of the Boskenna Bay, whereas the nine lots of the libelant’s fruit, which are compared with other sales of the same brands, form about one-sixth part,of all the libelant’s goods. The latter are, apparently, therefore, a better guide, because a larger proportion and a part of the very goods injured. As these goods were moreover for the same consignee, they were probably more nearly together, and hence somewhat similarly affected, while the other parts of the cargo were probably more unequally affected.

I cannot give much consideration to the general statement that this cargo was a “banner” cargo, since no equivalent in money value of the banner quality is stated. The prices realized upon the sale do not afford much support to this general testimony, nor has the libelant shown that the purchasers at the sale understood, as a rule, that they were buying damaged goods. In the absence of direct proof of the value of fruit of the same brand as the libelant’s in this market, in a sound condition on the day of delivery, the most satisfactory mode of ascertaining the presumed amount of injury to the libelant’s goods from frost would have been to compare the price that the goods actually brought in their damaged condition with the prices brought, within a wegk before or after, by other brands of the same invoice value as the libelant’s goods. The Colon, supra. If this was not obtainable, then proof of the value abroad would have been competent, with additions for differences in market. So a comparison of the invoice values of all the different lots of the libelant’s goods among themselves, when compared with the prices brought on the sales of nine similar brands of fruit in a sound condition, would have furnished, it would seem, pretty trustworthy evidence of the sound value of the 46 lots. The differences in the invoice prices might fairly be taken to correspond with the differences in value here; if not exactly, yet approximately, in the absence of all direct evidence of value. None of this proof was introduced. The average values of all the Palermo fruit at that time is only reasonable as a last resort, when all other means of proof were exhausted. The Rossend Castle, supra; The Scotland, 105 U. S. 35, 36.

*615But I must regard even this as more trustworthy than the extremely loose, vague, and widely discrepant estimates of persons who knew nothing of the actual injury to the fruit, and who gave estimates of what injury might happen to fruit exposed for 12 hours to a certain temperature. In the absence of any better testimony, I take, therefore, the average values of Palermo fruit during the week before and after the arrival of the Boskonna Bay, and compute the damage to the 46 lots by a comparison of tiróse average values, as the sound price, with the prices realized on the sale of the 46 lots in their damaged condition. The damage upon the nine lots will be taken at the difference between the amount at which they sold and the average sales of the similar nine lots in evidence. The parties will probably be able to make up the amount due upon the above basis without further reference.