Owen, J.

The policy insures plaintiff against damage resulting to his automobile “by being in accidental collision during the period insured with any other automobile, vehicle, or object.” Judgment was rendered in the lower court on the theory that the forcible contact of the automobile with the street upon the occasion of the upset constituted a collision within the meaning of the policy provision referred to. The appellant contends that the contact resulting from the upset is not a collision within the meaning of the terms of the policy, and that is the sole question before us.

It was held in Wettengel v. United States “Lloyds,” 157 Wis. 433, 147 N. W. 360, that the language of this policy provision- did not cover damages resulting to an automobile by its running off the main road and down a bank into a river. In that case the doctrine ejusdem generis was applied to the words “automobile, vehicle, or object,” and it was held that to entitle plaintiff to recover the collision must have occurred with another automobile, vehicle, or some similar objéct. If the doctrine of that'case is to be followed the judgment must be for the' defendant. It is urged by the appellant that the application of the doctrine ejusdem generis was unnecessary to the conclusion there reached, and that it was inadvertently applied and has no proper application to'the words as used in the policy *535before us. It is pointed out that the doctrine of ejusdem generis does not apply when the specific words embrace all objects of their class so that the general words must bear a different meaning from the specific words or be meaningless; that an automobile is a vehicle, and the term “vehicle” includes everything in which persons or things can be carried or transported, hence the words “automobile” and “vehicle” embrace all objects of their class, and the word “object” means a different kind of an object or it means nothing.

By the rule of construction known as ejusdem generis, 'general words following particular words are limited to other species of the same genus. “The particular words are presumed to describe certain species and the general words to be used for the purpose of including other species of the same gemís. The rule is based on the obvious reason that if the legislature had intended the general words to be used in their unrestricted sense they would have made no mention of the particular classes.” 36 Cyc. 1120. It has been held that the rule does not apply where the specific words embrace all objects of their class so that the general words must bear a different meaning from the specific words or be meaningless. United States C. Co. v. Cooper, 172 Ind. 599, 88 N. E. 69. We think the reason supporting the rule also dictates the exception, and that the exception applies to the words of this policy provision. Unless the word “object” as here used be construed as including an object of a different class it is meaningless, as the term “vehicle,” it seems to us, includes every species within the genus. We are disposed to construe this provision as sufficiently broad to include a collision with objects other than automobiles or vehicles, and withdraw the contrary intimation made in Wettengel v. United States "Lloyds,” supra. This requires us to determine whether the forcible contact of the automobile with the ground, as a result of the upset, constitutes a collision.

*536With the definitions of lexicographers as a basis, it is easy to demonstrate that the incident resulting in damage to plaintiff’s automobile constituted a collision. Thus, “A collision is the ‘meeting and mutual striking or clashing of two or more moving bodies, or of a moving body with a stationary one.’ Century Diet. ‘Object’ is defined to be ‘that which is put, or which may be regarded as put, in the way of some of the senses; something visible or tangible.’ Webster, Diet. An automobile is an object. Upon the overturning of an automobile its forcible contact with the earth constitutes a ‘mutual striking or clashing of a moving body with a stationary one.’ Hence the forcible contact of the automobile with the earth on the occasion of the upset constituted a collision.” Upon its face this appears to be good logic, but the conclusion is neither convincing nor satisfying. One instinctively withholds assent to the result. The reason is that it makes a novel and unusual use and application of the word “collision.” We do not speak of falling bodies as colliding with the earth. In common parlance the apple falls to the ground, it does not collide with the earth. So with all falling bodies. We speak of the descent as a fall, not a collision. In popular understanding a collision does not result, we think, from the force of gravity alone. Such an application of the term lacks the support of “widespread and frequent usage.”

While it is true that insurance contracts should be construed most strongly against the insurer (French v. Fidelity & C. Co. 135 Wis. 259, 115 N. W. 869; Kelly v. Fidelity Mut. L. Ins. Co. 169 Wis. 274, 172 N. W. 152), yet they are subject to the same rules of construction applied.to the language of any other contract. It is a fundamental rule that the language of a contract is to be accorded its popular and usual significance. It is not permissible to impute an unusual meaning to language used in a contract of insurance any more than to the language of any other contract. The incident causing the damage to the automobile here in *537question is spoken of in common parlance as an upset or tip-over. If it were the purpose to insure against damage resulting from such an incident, why should not such words, or words of similar import, have been used? We cannot presume that the parties to the contract intended that an upset should be construed as a collision in the absence of a closer association of the two incidents in popular understanding.

The term “collision” has not received frequent consideration by courts. A number of cases in which the term has been considered have been collated by the industry of counsel and cited to our attention. We here preserve a reference to the following for the future convenience of bench and bar: Newtown Creek T. Co. v. Ætna Ins. Co. 163 N. Y. 114, 57 N. E. 302; London Assurance v. Companhia De Moagens Do Barreiro, 167 U. S. 149, 17 Sup. Ct. 785; Cline v. Western Assur. Co. 101 Va. 496, 44 S. E. 700; Harris v. American C. Co. 83 N. J. Law, 641, 85 Atl. 194; Hardenbergh v. Employers’ L. A. Corp. 80 Misc. 522, 141 N. Y. Supp. 502; Stuht v. United States F. & G. Co. 89 Wash. 93, 154 Pac. 137; O’Leary v. St. Paul F. & M. Ins. Co. (Tex.) 196 S. W. 575; Graham v. Ins. Co. of N. A. 220 Mass. 230, 107 N. E. 915. None of these cases, however, deal with a situation similar to the one here presented, unless it be Harris v. American C. Co.., supra. That case was referred to in Wettengel v. United States “Lloyds,” 157 Wis. 433, 147 N. W. 360, where it was said that it had but advisory value in this court and its soundness was doubted. A further consideration of the subject does not dispel the doubts there expressed.

For the purpose of showing a practical construction of the contract on the part of the company, the plaintiff proved by the agent who delivered the policy in question that the company provided, and SQmetimes used, another form covering the damage resulting from collision, which specifically excluded “damage caused by striking any portion of the *538roadbed or by striking the rails or ties of street, steam, or electric railroads.” It is argued that because in some instances the company used a form specifically excluding damage caused by striking any portion of the roadbed, or by striking the rails or ties of street, steam, or electric railroads, a purpose is indicated to assume responsibility by the use of the one form for damage excluded by the express terms of the other. Without stating whether, in our opinion, this testimony proves, or tends to prove, a practical construction of the contract here in question by the company, we are clear that it does not tend to prove a practical construction in harmony with that for which .respondent contends. As we construe the clause quoted, it has no reference to damage caused by upsets, but it does exclude damage caused by projecting portions of the roadbed in the course of travel. It is common knowledge that an automobile traveling along a highway frequently strikes an unevenness of surface in the roadbed sufficient to do damage to the automobile, and this, we think, is the damage excluded by the clause in the other form sometimes used by the company. Such circumstance was held not to constitute a collision in Dougherty v. Ins. Co. of N. A. 38 Pa. Co. Ct. Rep. 119, and it seems not •improbable that the form exempting such liability is sometimes used by defendant to avoid just such contentions as were there made. We do not think the introduction of the second form throws any light on the proper construction of the contract before us.

We are of the opinion that the damages sustained were not the result of a collision, and that the judgment should be reversed.

By the Court. — Judgment reversed, and cause remanded with directions to enter judgment dismissing the plaintiff’s complaint.