74 A Study of Student Loans and about pressure to force payment. It therefore causes the delinquent to sit up and take notice and brings him to the full realization of his duty to his social group and in turn to society. It is believed by the proponents of the practice of the group guarantee that two or three per cent. premium is sufficient to protect the principal of the fund. The Guarantee Fund As may be found in Appendices A and B, the guarantee fund isderived from a surcharge to students who borrow. When the several participants of the group have all paid, as under the Harmon plan, a refund is made deducting losses and paying six per cent. on the funds which have been paid in as a surcharge. Under the German plan every Student is made to contribute to the guarantee fund. The objection that can be raised to the Harmon plan is that the group is too small. In case of a default or accident to one or more members of the group the other members are too severely punished. It is contrary to sound principles for the distribution of losses. However, cognizance must be taken of the fact that an alterna tive is difficult to find while the funds are so limited. As more funds are loaned the risk should be spread over a larger number of individuals. This can come only with time. Under the German plan, on the other hand, we find the opposite extreme. All students are made to pay into the guarantee fund whether they borrow or not. This is no doubt unwise, since it places the bürden of loss on those outside the borrowing group and should be objectionable even to the borrowers. There is no doubt but that the principles underlying the group guar antee plan are sound. This is in harmony with social institutions and human activity. All losses must be borne by some one and the sound method should be to have a group, community, or generation bear its own losses when such losses are the result of its own activities. The objection to be levied against the plan is, therefore, not against the group guarantee itself, but against the grouping for such guarantee and the selection of the risk. Without attempting to show the faults of the present grouping and selection of individuals, it should be in order to discuss a more favorable form, and thus make self-evident the shortcomings of the present grouping and selection. To obtain the best results it would seem that a grouping by classes within the College would be the best form. Use the dass as a group. This should prove effective because there is already existing a certain bond in the dass. The members of the same dass spend four years in College together and perpetuate this grouping in their alumni association.