(dissenting).

I dissent. I think that the panel should grant a rehearing. As a member of the panel that first heard the case, I concurred in Judge Madden’s opinion. However, the petition for rehearing and the decision in Reed v. Reed, 1971, 404 U.S. 71 , 92 S.Ct. 251 , 30 L.Ed.2d 225 , persuade me that a rehearing should be granted, and the judgment appealed from reversed.

' In Reed v. Reed, supra, a unanimous Court in an opinion by the Chief Justice, held that a state statute giving preference to males over females in the administration of estates violates the equal protection clause of the Fourteenth Amendment. In so doing, the Court held that the difference-in sex of the competing applicants for letters of administration bore no “rational relationship to a state objective that is sought .to be advanced.” (92 S.Ct. p. 254.) The principle of the Reed decision is applicable here because federal action of a kind which, if done by a state, would violate the equal protection clause of the Fourteenth Amendment, is a violation of • the due process clause of the Fifth Amendment. Bolling v. Sharpe, 1954, 347 U.S. 497 , 74 S.Ct. 693 , 98 L.Ed. 884 .

It is not yet clear whether classification based upon sex is “suspect,” and therefore can be sustained only upon a showing by the government that a compelling interest makes the classification necessary. Cf. Shapiro v. Thompson, 1969, 394 U.S. 618 , 89 S.Ct. 1322 , 22 L.Ed.2d 600 (classification based on length of residence as affecting right of interstate movement); Loving v. Virginia, 1967, 388 U.S. 1 , 87 S.Ct. 1817 , 18 L.Ed.2d 1010 , and Korematsu v. United States, 1944, 323 U.S. 214 , 65 S.Ct. 193 , 89 L.Ed. 194 (classification based on race); Sail’er Inn, Inc. v. Kirby, 1971, 5 Cal.3d 1, 17-18 , 95 Cal.Rptr. 329, 340 , 485 P.2d 529, 539-540 (classification based on sex). I think, however, that eventually the Supreme Court will so hold. Be that as it may, Reed, supra, does not purport to go that far, but the criterion that it does establish requires that we hold that the regulation here in question is invalid. I can find no rational relationship between the regulation and any legitimate objective that it can be said to seek to achieve.

It is not enough to say that there are physical differences between men and women, or that men and women are not fungible. Rather, the question is, does the regulation take account of the differences in a rational way? Nor is it enough to say that only women can become pregnant and that therefore it is rational to treat pregnancy differently from all other physical disabilities. There are other disabilities to which only females are subject, just as there are some to which only men are subject. For example, no woman can suffer from “undiagosed enlargement or mass of the testicle or epiditymis,” which under Army Reg. 40-501, ff 2-14 m(2), disqualifies a man for military service. The question is, in view of the general obligation to treat the sexes alike established in Reed, supra, whether there is a rational basis for the special and unique treatment of pregnancy embodied in the regulation. I can find none. Indeed, the regulation seems to me to be quite irrational.

I start with certain assumptions. One is that the government has a legitimate interest in the health and welfare of pregnant women and their unborn children in general, and in particular of those women who are federal employees. A second is that the government (including especially the military branch of government) also has a legitimate interest *1379 in maintaining effective military forces. If there be any other interest that could be advanced as the basis for promulgating Regulation 36-12, I am not aware of it.

I first consider subsection (a) of Regulation 36-12. It creates, among others, the following classifications: (1) pregnant female officers versus non-pregnant female officers; (2) pregnant female officers versus non-pregnant female employees who are not officers; (3) detectable pregnancy versus all other physical conditions that may or can endure for not more than nine months in the persons of male or female officers. Do such classifications have any rational relationship to the health or welfare of the pregnant female officer or her unborn child, or both, as a separate class? I can find none. Even if every female government employee were included in the class, it could not stand. To sustain it, there would have to be some evidence that continued employment of such female employees would be detrimental to their health. I am aware of no such evidence. On the contrary, centuries of human experience attest to the reality that pregnant women work throughout their pregnancies, whether that work be performed in fields, mills, factories, stores or offices, or in homes cleaning house, cooking, and tending their husbands and their other children. Of course, not all pregnant women work and at least some pregnant women who work have health problems. But neither the fact of work nor the fact that one is paid for work has any necessary relationship to maternal and infant health.

No doubt it is true that the particular work that a pregnant woman may be called on to perform may affect her health. So, too, the time during her pregnancy when she is called upon to perform it may have such effect. But these considerations do not save subsection (a). That subsection takes no account of them; it cannot be sustained on the basis of criteria that it does not fit.

Similarly, I can find no rational relationship between subsection (a) and the interest of the government, through the Air Force, in maintaining effective personnel. Is there any evidence that detectably pregnant officers are less able to perform their duties than non-pregnant female employees with similar jobs? I pose this question because the sole criterion for discharge that the regulation establishes is medically detected pregnancy. One whose pregnancy is detected at eight days is treated differently from all other non-pregnant female employees. The apppropiaté inquiry, therefore, is whether or not women at any detectable stage of pregnancy are measurably less able to perform their tasks than non-pregnant women employees. Again, I am unaware of any empirical data that support that affirmative.

It is immaterial whether a female who is, say, eight months pregnant, may be less able to work than a non-pregnant female. It is equally immaterial whether a pregnant female is less able to work under combat conditions. The regulation does not use these criteria; it totally disregards them.

Of especial constitutional significance is the classification of detectable pregnancy versus any other kind of temporary “physical disability.” The fact alone of detectable pregnancy mandates discharge. No other temporary physical condition results' in mandatory discharge. 'There is obviously a difference between pregnancy and all other temporary physical conditions. But under Reed, supra, a classification based on that difference is not constitutional unless the difference bears some rational relation to the governmental interest served. Is there any evidence that pregnancy has some effect on ability to function as an officer that is different from any other temporary physical condition? For example, is there any reason to believe that a female officer who has suffered a fractured leg is better able to perform her job than a female officer who is eight days pregnant? The former gets medical leave and retains her commission; the latter is discharged. Why? If this be rational, nothing is irrational !

*1380 It is no answer at all to say, as does the court’s opinion, that “other personnel . . . might [be] disabled and made useless. Those events would have been the fortunes of war.” Under subsection (a) it is completely immaterial whether the pregnant officer is in combat or not. It is also completely immaterial how or when any other “disability” occurred. An officer who fractures her leg skiing is not subject to mandatory discharge any more than one whose leg is fractured in combat. Moreover, under subsection (a) any event that occurs after detection of pregnancy is immaterial. The officer is discharged “with the least practical delay.” She is no less dischargeable if she miscarries three days later.

The conclusion follows that subsection (a) of the regulation is unconstitutional on its face.

I turn to subsection (b) (d). Here discharge is mandated upon proof that a female officer has given birth to a living child while she was an officer. No other governmental interests appear here than those mentioned in the discussion of subsection (a).

Here is a non-exclusive list of classifications created by subsection (b) (d): (1) commissioned officers who are mothers versus commissioned officers who are fathers; (2) commissioned female officers who give up their children for adoption versus commissioned male officers who give up their children for adoption; (3) commissioned female officers whose infants have been voluntarily or involuntarily aborted versus commissioned female officers who delivered their infants; (4) commissioned female officers who have babies versus other female employees of the federal government who have babies.

Obviously the sexes are not fungible. That fact, however, does not show that the difference between the persons in each of the above classifications has “a fair and substantial relation to the object of the legislation” { Reed, supra). What and where is the compelling interest of the government that is necessarily served by these classifications ? Why should a female officer whose infant is adopted lose her commission (Captain Struck’s situation), and a male officer whose infant is adopted keep his? Why should a female officer who aborts her child keep her commission, and a female officer who delivers her child forfeit her commission? Why should a female officer who has a baby forfeit her career, and every other female federal employee keep hers if she wishes to do so ? I cannot find an answer to these questions, and I therefore conclude that subsection (b) (d) is likewise unconstitutional on its face.

To grant relief here is not to transfer to judges the task of running the army. (See Orloff v. Willoughby, 1953, 345 U.S. 83, 93 , 73 S.Ct. 534 , 97 L.Ed. 842 ). The Armed Forces are not above the Constitution, but are subordinate to it. If they adopt regulations that violate the Constitution, it is the plain duty of the courts to say so. I cannot believe that any of my colleagues would hesitate for a moment to strike down a regulation stating that no person of African ancestry can be commissioned, or one stating that all such persons now holding commissions shall be discharged. In principle, this case is no different.

I would grant a rehearing and reverse the judgment.