The Orville has Mochlans, an (almost) all-male species. They reproduce in some unexplained method that produces a fertilized egg.

The issue here is that if one of the mates lays an egg and the other fertilizes it, they are functionally hermaphrodites and the male designation basically only implies a certain set of genitalia (and perhaps hormones) rather than reproductive function. For this to be a functionally male-only species, there would need to be some other mechanism that produces the egg.

For the purposes of this question, femaleness is defined as incubating a small number of large cells to be fertilized by a mate (i.e. production of eggs), whereas maleness is defined as producing a large number of small cells (sperm) that compete to fertilize an egg. Other traits like lactation and physical appearance are irrelevant.

One mechanism I can think of involves ejaculating sperm into a common pool, but there are a few problems with this:

• What would prevent the creation of single-parent offspring? It needs to be sexual reproduction, not a sort of male parthenogenesis.


• What would prevent the conception of millions of offspring? Ideally, this should produce one offspring most of the time for a humanoid.

This mechanism should also be natural since it's conceivable that technology could be used to create artificial eggs from stem cells and an artificial incubator could be used to simulate pregnancy (this is less practical than an artificially all-female species, but should be possible as far as I know). This should be something that could occur naturally through evolution.

EDIT: as comments have pointed out, you can't exactly have "males" for a single-sex species because "male" implies the existence of "female" (The designation technically works for Mochlans since females are just incredibly rare), so I'll refine this down to a species that reproduces through some mechanism exchanging large quantities of motile gametes (in the millions, like sperm) to produce a small number of offspring (usually 1, sometimes more, like humans). Gametes are the same size from both mates, so it is a single-sex species that reproduces sexually. In addition, neither mate has organs for incubating offspring like you might expect from mammals. Although an egg might be produced through this process, "laying" them also isn't possible, so the egg must be grown externally. Essentially, neither mate can perform what might be understood as a "female" role, other than incubating eggs (whether this means sitting on them or keeping them in a pouch.) or lactation since those aren't, strictly speaking, female-exclusive. (Emperor Penguins and Seahorses are good examples for male egg-bearers and even human men have been known to lactate in some rare cases)

Basically the idea here is to make the method of reproduction look like something that we as humans would observe and decide is most simply explained as an "all male" species.

Yes. And as you have supposed, simply ejecting "sperm" into a common pool (or more technically, motile isogametes) is the most straightforward way to do it. How do you prevent single-parent offspring? The same way that many real-world isogamous species do: by instituting mating types. I.e., you may have 7, 11, 29, or even larger (not necessarily prime) numbers of distinct biological sexes distinguished by nothing more than the chemical markers on their gametes that prevent them from fertilizing themselves, and with no more sociological significance than blood types.

How do you prevent it from resulting in millions of offspring at a time? Well, that's easy--you just decide to be careful about what counts as "offspring". If you count individual embryos, that might be tough--but externally fertilized individual embryos don't have a fantastic survival rate if they aren't intentionally cared for, and that's hard to do when they're microscopic! So, just establish a convention that offspring don't count until they reach a certain minimum developmental stage--anything that dies earlier is just a "failed pregnancy". Then you just need to figure out how to kill off most of them--which, again, isn't hard. If you want to make sure that there is a natural tendency for any particular intentional procreative event to average one surviving offspring, just make the larva cannibalistic--only one or two will manage to make it past "fetus" stage after having eaten all of their brothers.

For a more complicated option, consider hybridization with sexual parasitism--in which case you really can have "males" of one species, as distinguished from the males and females of a different species. The simplest way to work this is to have the all-male species' sperm hijack the eggs of their host species' females, ejecting the host's DNA and resulting in the incubation and birth of a new male every time, with no (or at least minimal--perhaps they preserve the equivalent of mitochondrial DNA from the host mother) gene transfer between species. While mechanically that looks like sexual reproduction, however, genetically speaking it is effectively asexual, with all males being nuclear clones of each other. There are a couple of ways around this.

1. Maternal DNA is not discarded during fertilization, but during
   meiosis. I.e., every male is actually a hybrid, carrying the unique
   DNA of his own species from his father, and half of the DNA from his
   extra-specific mother. When he manufactures sperm, however, the
   maternal DNA is preferentially discarded, ensuring that the paternal
   DNA is propagated unchanged between generations, and merely
   re-hybridized with new host females each time. This ensure no
   permanent gene flow between the populations of the all-male
   parasites and their female-having host species, and limits evolution
   of the parasite male chromosomes to spontaneous mutation like an
   asexual species, but still allows the parasite male population to
   effectively evolve alongside, and benefit from the evolution of,
   their host population.



2. 
   

   If we want gene transfer between males, you will need to
   introduce an additional step--a host species which is internally
   fertile with both males and females, a hybrid form which may or may
   not be internally fertile, and the parasitic all-male species. This
   can go a few different ways, depending on the fertility
   characteristics of the hybrid form(s), but the next simplest option
   in this case is to make male hybrids either completely sterile or
   simply non-existent (i.e., any hybrid male embryos are simply
   non-viable, like a human YY embryo would be), leaving only female
   hybrids able to procreate with at least the all-male parasite species,
   and optionally with the the host disexual species.
   In either case, we assume that there
   is some unique genetic material (i.e., mis-matched chromosomes) that
   can be distinguished between species and does not transfer between
   them, although there may be some mixing in other parts of the
   genome--this shouldn't be too big of a deal, since the two species
   involved would have to have started out pretty closely related
   anyway in order for hybridization to be possible, so they'll only be
   sharing genes that they already had in common anyway. Parasite
   male-species reproduction then occurs in two steps:

   
   
   

   a) a parasite male mates with a host female to produce a hybrid female

   
   
   

   b) a parasite male mates with a hybrid female, resulting in a parasite male who is the nearly-pure genetic child of his father and maternal grandfather.

   
   
   

   Of course, we have to explain why hybrid females can't produce more hybrid females as offspring when mating with parasite males; if they could, we'd be left to wonder why this does not result in simply discarding the host species and adopting the hybrids as the new females of the formerly-parasitic species. Once again, we can go in many directions with this, but we'll pick the simplest--as in option 1, hybrid meiosis discards the identifiably-maternal genome, so hybrids can in fact only produce parasite male offspring, not new hybrid females after all. This results in fully-recombinant sexual reproduction between members of the all-male species, spread out over an intermediate generation; and while unlike option 1 it does permit some gene flow between parasite and host species, that can be limited to only a fraction of the chromosome complement of each species, which would have to remain similar anyway for hybridization to remain viable, and does not impact the purity of the distinctive male genetic line or allow the distinctive genes of the parasite species to flow into the host species population.

   
   

For a (somewhat more complicated) real-life example of this kind of all-male species that reproduces through hybridization consider a certain clade of Australian fish in a four-way species complex. Now, not 100% of them are male, but it's pretty dang close.

Now, why is this two-or-three-part species complex not just considered one species with a bunch of different gender morphs, regardless of their ancestry from two originally separate species? Aside from the genetic arguments (which would not necessarily be readily available to them or to humans at first glance), it is not difficult to come up with cultural ones. A fairly straightforward explanation, for example, could be that the host species is just not that smart--equivalent to, say, bonobos, while the parasite species is on-par with modern humans. This is analogous to the situation with Larry Niven's Peirson's Puppeteers--their sexual host species (with whom they have zero genetic transfer, as it requires two Puppeteers to inseminate a host female, completely ignoring the female's contribution) is essentially livestock. If they don't want to consider their sexual hosts to be of the same species as them, and the hosts can't exactly protest that categorization, humans are likely to along with it.

Oglaf, a totally NSFW comic which I won't link to, has an all male tribe. It's mostly a joke though, with men impregnating men and babies magically coming out of their assets.

For a purely male species with the requirements you provided, you could have it like this:

• Individuals reproduce by binary fission;


• If an individual has received sperm from another individual, the receiver can mix their own original genes with those they received; and not in a small group of special cells, but throughout their whole body.

Some bacteria like our symbiont E. coli do something similar, it's called conjugation and goes like this:

It wouldn't be a stretch to adapt this to multicellular life in a work of fiction.

Virii also seem to match what you want. The absolute vast majority of them only ever impregnates other life with their DNA; only a small minority may be impregnated, and always by virii of other species. Now, granted, they are acellular; but that, again, can be adapted to multicellular life.

Last thing I can think of are hybrids of any two species, for which only the male hybrids are viable. They don't belong to either of their parents' species, and the whole population will be male. What allows them to reproduce is mating with a female of either parents' species; the amount of viable offspring may be small, producing one might be like winning the lottery. This would keep their numbers really low. Add that for some reason the male cares for the egg, which is not unusual in nature, and there you have it.

Haldane's Rule says unlikely (but doesn't always hold, even on Earth.)

Back in 1922, a smart guy named Haldane formulated a rule: if, in a species hybrid, only one sex is inviable or sterile, that sex is more likely to be the heterogametic sex. (The heterogametic sex is the one with two different sex chromosomes; in therian mammals, for example, this is the male.)

Or you could read up on an all-female fish species if that also interests you.

The eggs are ancient.

In this scenario, the males fertilize eggs in the manner of fish or amphibians, spraying milt on one or more eggs and so triggering their development.

The eggs in question come from a huge secret cache, a shrine of sorts to this species. In this cave are hundreds of thousands of unfertilized eggs, left there by the long vanished females. The males tend these unfertilized eggs and keep them viable. When reproduction is in order a male will take one out, fertilize it and tend it much in the way a male fish like a bass or a tilapia will tend his nest of fertilized eggs.

There will never be any more eggs. When they run out the species goes extinct. But that will not be for a long time - there are a lot of eggs in this cave.

Since Female tends to imply bearing young, no unless the males impregnate some other host that will then bear young. In this case the race of all males is more like a race of parasites.

The trick here is that the second race must be able to mate viably with itself (has males and females and can make babies). That keeps that other race as a separate race whose biological systems are hijacked by the parasitical race.

Extreme sexual dimorphism

You can have a "female" organism that has evolved into a sessile, predatory form, and a "male" organism that is motile.

The female produces ova, and also secretes pheromones that attract mature males (as well as other prey). Similar to the praying mantis, if the female is well fed enough it won't kill the male. With the development of male sentience, this is pretty easy to accomplish. Older and weaker males get absorbed, similar to what happens to anglerfish.

Being a top predator and having no natural enemies, the female organism never developed intelligence and is optimized for reproduction. Very young females are motile, and can wander far enough from the mother to establish a new lair. Males, on the other hand, can go wherever they want.

Some fishes have rudimentary males. Rudimentary females are not known, but can be imagined. If males are something as sea horses, with eggs growing in the pocket on the male's body with females very small and living also on the males, in the same pocket, only forever.